From 17c2d2df5ec413f0f3987a3b60576e034df24bbf Mon Sep 17 00:00:00 2001 From: "github-actions[bot]" <41898282+github-actions[bot]@users.noreply.github.com> Date: Thu, 12 Sep 2024 06:59:39 +0000 Subject: [PATCH] Add changes --- data.js | 1569 +++++++++++++++++++++++++++------------------------- skips.json | 392 ++++++------- 2 files changed, 1014 insertions(+), 947 deletions(-) diff --git a/data.js b/data.js index 68e7f65..6190811 100644 --- a/data.js +++ b/data.js @@ -1,5 +1,369 @@ var data = [ + { + "config_readme": "\u003cp\u003eDescribe how to configure the workflow (using config.yaml and maybe additional files).\nAll of them need to be present with example entries inside of the config folder.\u003c/p\u003e\n", + "data_format": 2, + "description": null, + "formatting": "[DEBUG] \n[DEBUG] \n[DEBUG] In file \"/tmp/tmpowss2jy5/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[DEBUG] \n[DEBUG] \n[INFO] 1 file(s) would be changed \ud83d\ude2c\n[INFO] 4 file(s) would be left unchanged \ud83c\udf89\n\nsnakefmt version: 0.10.2", + "full_name": "loganylchen/snakemake-epitranscriptome", + "latest_release": null, + "linting": null, + "mandatory_flags": { + "desc": null, + "flags": null + }, + "report": true, + "software_stack_deployment": { + "conda": true, + "singularity": false, + "singularity+conda": false + }, + "standardized": true, + "stargazers_count": 0, + "subscribers_count": 1, + "topics": [], + "updated_at": 1726116394.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "A Snakemake pipeline to analyze RNA-seq expression data", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp31et65mg/workflow/rules/diffexp.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp31et65mg/workflow/rules/quantification.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmp31et65mg/workflow/Snakefile\": Keyword \"input\" at line 44 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmp31et65mg/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp31et65mg/workflow/rules/plots.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp31et65mg/workflow/rules/splicing.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp31et65mg/workflow/rules/trim.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp31et65mg/workflow/rules/align.smk\": Formatted content is different from original\n[INFO] 7 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "kristinassong/RNAseq", + "latest_release": null, + "linting": "Lints for rule star_index (line 1, /tmp/tmp31et65mg/workflow/rules/align.smk):\n * Param idx is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n\nLints for rule star_align (line 27, /tmp/tmp31et65mg/workflow/rules/align.smk):\n * Param out_prefix is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n * Param idx is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n\nLints for rule genomecov (line 97, /tmp/tmp31et65mg/workflow/rules/align.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule kallisto_quant (line 38, /tmp/tmp31et65mg/workflow/rules/quantification.smk):\n * Param outdir is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n\nLints for rule tx2gene (line 67, /tmp/tmp31et65mg/workflow/rules/quantification.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule merge_kallisto_quant (line 80, /tmp/tmp31et65mg/workflow/rules/quantification.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule deseq2 (line 3, /tmp/tmp31et65mg/workflow/rules/diffexp.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Param kallisto_dir is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n * Param out_dir is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n\nLints for rule volcano_plot (line 22, /tmp/tmp31et65mg/workflow/rules/diffexp.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule pca (line 1, /tmp/tmp31et65mg/workflow/rules/plots.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule multiqc (line 17, /tmp/tmp31et65mg/workflow/rules/plots.smk):\n * Param outdir is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n\nLints for rule rmats (line 1, /tmp/tmp31et65mg/workflow/rules/splicing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule filter_rmats (line 23, /tmp/tmp31et65mg/workflow/rules/splicing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Param dir is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n\nLints for rule rmats_paired_env (line 46, /tmp/tmp31et65mg/workflow/rules/splicing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule rmats_paired (line 62, /tmp/tmp31et65mg/workflow/rules/splicing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule filter_rmats_paired (line 85, /tmp/tmp31et65mg/workflow/rules/splicing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Param dir is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 0, + "subscribers_count": 1, + "topics": [ + "rna-seq-pipeline", + "snakemake-workflow" + ], + "updated_at": 1726112847.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "Workflow for processing UAS imagery into data on bird location and species for near real-time monitoring in the Everglades", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpopwg9ani/weecology-everwatch-workflow-2e6fceb/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "weecology/everwatch-workflow", + "latest_release": "v0.2.0", + "linting": "WorkflowError in file /tmp/tmpopwg9ani/weecology-everwatch-workflow-2e6fceb/Snakefile, line 5:\nWorkflow defines configfile /blue/ewhite/everglades/everwatch-workflow/snakemake_config.yml but it is not present or accessible (full checked path: /blue/ewhite/everglades/everwatch-workflow/snakemake_config.yml).\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 0, + "subscribers_count": 4, + "topics": [], + "updated_at": 1726093707.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": null, + "formatting": "[DEBUG] \n[WARNING] In file \"/tmp/tmpgdnlbxyx/workflow/Snakefile\": Keyword \"input\" at line 142 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmpgdnlbxyx/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpgdnlbxyx/workflow/rules/trim.smk\": Formatted content is different from original\n[DEBUG] \n[ERROR] In file \"/tmp/tmpgdnlbxyx/workflow/rules/test_trim.smk\": SyntaxError: L28: Unrecognised keyword \u0027ouput\u0027 in rule definition\n[DEBUG] In file \"/tmp/tmpgdnlbxyx/workflow/rules/test_trim.smk\": \n[DEBUG] In file \"/tmp/tmpgdnlbxyx/workflow/rules/human_filter.smk\": Formatted content is different from original\n[INFO] 1 file(s) raised parsing errors \ud83e\udd15\n[INFO] 3 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "Zjx01/RNA_pipeline_Snakemake", + "latest_release": null, + "linting": "FileNotFoundError in file /tmp/tmpgdnlbxyx/workflow/Snakefile, line 14:\n[Errno 2] No such file or directory: \u0027/Users/zhaoj11/Documents/RNA_pipeline_Snakemake/config/config.yaml\u0027\n File \"/tmp/tmpgdnlbxyx/workflow/Snakefile\", line 14, in \u003cmodule\u003e\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 0, + "subscribers_count": 1, + "topics": [], + "updated_at": 1726081445.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "Workflow to identify the centromere dip region (CDR) from methyl BAMs", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpumxp1g0a/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[INFO] 1 file(s) would be changed \ud83d\ude2c\n[INFO] 1 file(s) would be left unchanged \ud83c\udf89\n\nsnakefmt version: 0.10.2", + "full_name": "koisland/CDR-Finder", + "latest_release": null, + "linting": null, + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 0, + "subscribers_count": 1, + "topics": [], + "updated_at": 1726071971.0 + }, + { + "config_readme": "\u003cp\u003eDescribe how to configure the workflow (using config.yaml and maybe additional files).\nAll of them need to be present with example entries inside of the config folder.\u003c/p\u003e\n", + "data_format": 2, + "description": null, + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpbyphbz3g/workflow/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "vkpat/short-read-mapping-and-QC-snakemake-pipeline", + "latest_release": null, + "linting": "WorkflowError in file /tmp/tmpbyphbz3g/workflow/Snakefile, line 1:\nWorkflow defines configfile config.yaml but it is not present or accessible (full checked path: /tmp/tmpbyphbz3g/config.yaml).\n", + "mandatory_flags": { + "desc": null, + "flags": null + }, + "report": true, + "software_stack_deployment": { + "conda": true, + "singularity": false, + "singularity+conda": false + }, + "standardized": true, + "stargazers_count": 0, + "subscribers_count": 1, + "topics": [], + "updated_at": 1726070051.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": null, + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmptmii62qa/rules/renv.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmptmii62qa/Snakefile\": Formatted content is different from original\n[INFO] 2 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "deer-marketing-lab/dsms-lecture-ugc-ratings", + "latest_release": null, + "linting": "Lints for rule build_html (line 16, /tmp/tmptmii62qa/Snakefile):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable run_r from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable log_all from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule build_pdf (line 29, /tmp/tmptmii62qa/Snakefile):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable run_r from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable log_all from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule clean (line 48, /tmp/tmptmii62qa/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule renv_install (line 10, /tmp/tmptmii62qa/rules/renv.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable runR from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule renv_consent (line 15, /tmp/tmptmii62qa/rules/renv.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule renv_init (line 20, /tmp/tmptmii62qa/rules/renv.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule renv_snap (line 25, /tmp/tmptmii62qa/rules/renv.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule renv_restore (line 30, /tmp/tmptmii62qa/rules/renv.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 0, + "subscribers_count": 0, + "topics": [], + "updated_at": 1726069627.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "[INSERM U981] WGS\u0026WES Pipelines", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmppavg8cvv/workflow/rules/init_pipeline.smk\": Formatted content is different from original\n[DEBUG] \n[ERROR] In file \"/tmp/tmppavg8cvv/workflow/Snakefile\": InvalidPython: Black error:\n```\nCannot parse: 62:0: \u003cline number missing in source\u003e\n```\n\n[INFO] In file \"/tmp/tmppavg8cvv/workflow/Snakefile\": 1 file(s) raised parsing errors \ud83e\udd15\n[INFO] In file \"/tmp/tmppavg8cvv/workflow/Snakefile\": 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "jinxin-wang/Genome_Sequencing_Analysis", + "latest_release": null, + "linting": "FileNotFoundError in file /tmp/tmppavg8cvv/workflow/rules/init_pipeline.smk, line 118:\n[Errno 2] No such file or directory: \u0027config/variant_call_list_TvN.tsv\u0027\n File \"/tmp/tmppavg8cvv/workflow/rules/init_pipeline.smk\", line 118, in \u003cmodule\u003e\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 0, + "subscribers_count": 2, + "topics": [], + "updated_at": 1726062684.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "Snakemake workflow to re-batch POD5 files by pore/channel", + "formatting": "[DEBUG] \n[ERROR] In file \"/tmp/tmp_ei86lq4/workflow/Snakefile\": InvalidPython: Black error:\n```\nCannot parse: 68:0: res = defaultdict(set)\n```\n\n[INFO] In file \"/tmp/tmp_ei86lq4/workflow/Snakefile\": 1 file(s) raised parsing errors \ud83e\udd15\n\nsnakefmt version: 0.10.2", + "full_name": "tbooth/pod5_by_pore", + "latest_release": null, + "linting": "Lints for snakefile /tmp/tmp_ei86lq4/workflow/Snakefile:\n * Absolute path \"/{p5}.pod5\" in line 17:\n Do not define absolute paths inside of the workflow, since this renders\n your workflow irreproducible on other machines. Use path relative to the\n working directory instead, or make the path configurable via a config\n file.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/configuration.html#configuration\n * Absolute path \"/{foo}.pod5\" in line 41:\n Do not define absolute paths inside of the workflow, since this renders\n your workflow irreproducible on other machines. Use path relative to the\n working directory instead, or make the path configurable via a config\n file.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/configuration.html#configuration\n * Mixed rules and functions in same snakefile.:\n Small one-liner functions used only once should be defined as lambda\n expressions. Other functions should be collected in a common module, e.g.\n \u0027rules/common.smk\u0027. This makes the workflow steps more readable.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/modularization.html#includes\n * Path composition with \u0027+\u0027 in line 17:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 41:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n\nLints for rule get_channels_in_pod (line 39, /tmp/tmp_ei86lq4/workflow/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule batched_channels_in_pod (line 46, /tmp/tmp_ei86lq4/workflow/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule get_channels_in_use (line 61, /tmp/tmp_ei86lq4/workflow/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 0, + "subscribers_count": 1, + "topics": [], + "updated_at": 1726061158.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": null, + "formatting": "[DEBUG] \n\u003cunknown\u003e:14: SyntaxWarning: invalid escape sequence \u0027\\-\u0027\n[DEBUG] In file \"/tmp/tmpdyuhxgko/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "fredjarlier/ProjectCNN", + "latest_release": null, + "linting": "/tmp/tmpdyuhxgko/Snakefile:78: SyntaxWarning: invalid escape sequence \u0027\\-\u0027\n \"{yes}.bam\"\n/tmp/tmpdyuhxgko/Snakefile:85: SyntaxWarning: invalid escape sequence \u0027\\-\u0027\n input:\nKeyError in file /tmp/tmpdyuhxgko/Snakefile, line 31:\n\u0027lanes\u0027\n File \"/tmp/tmpdyuhxgko/Snakefile\", line 31, in \u003cmodule\u003e\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 0, + "subscribers_count": 2, + "topics": [], + "updated_at": 1726058223.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "Mu-seq Workflow Utility ", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/insertion_identification.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/Snakefile\": Keyword \"input\" at line 93 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/Snakefile\": Keyword \"input\" at line 129 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/Snakefile\": Keyword \"input\" at line 168 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/Snakefile\": Keyword \"input\" at line 200 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[ERROR] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/ref.smk\": InvalidPython: Black error:\n```\nCannot parse: 18:4: run:\n```\n\n[DEBUG] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/ref.smk\": \n[ERROR] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/trimming.smk\": InvalidPython: Black error:\n```\nCannot parse: 45:4: log:\n```\n\n[DEBUG] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/trimming.smk\": \n[DEBUG] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/common_utils.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/quality_control.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/insertion_annotation.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/ref_utils.smk\": Formatted content is different from original\n[DEBUG] \n[ERROR] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk\": InvalidPython: Black error:\n```\nCannot parse: 14:7: output:\n```\n\n[DEBUG] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk\": \n[DEBUG] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/indexes_bam2sam.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/duplicate_removal.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/common.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/mapping.smk\": Inline-formatted keyword \"threads\" at line 80 had its comments relocated above it.\nPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/mapping.smk\": Formatted content is different from original\n[INFO] 3 file(s) raised parsing errors \ud83e\udd15\n[INFO] 10 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "tgstoecker/MuWU", + "latest_release": "v1.5.0", + "linting": "/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/ref.smk:129: SyntaxWarning: invalid escape sequence \u0027\\.\u0027\n \"logs/annotation_table/annotation_table.log\"\n/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/ref.smk:142: SyntaxWarning: invalid escape sequence \u0027\\.\u0027\n \".1.bt2\", \".2.bt2\", \".3.bt2\", \".4.bt2\", \".rev.1.bt2\", \".rev.2.bt2\",\n/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk:45: SyntaxWarning: invalid escape sequence \u0027\\s\u0027\n sed -i \u00271i Name\\\\tStrand\\\\tType\u0027 {output}\n/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk:53: SyntaxWarning: invalid escape sequence \u0027\\s\u0027\n read_typing=expand(\"results/te_typing/pre_sorting/{sample}/{sample}_te_types_merged.tsv\", sample=SAMPLES),\n/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk:510: SyntaxWarning: invalid escape sequence \u0027\\(\u0027\n/tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk:519: SyntaxWarning: invalid escape sequence \u0027\\(\u0027\nLints for rule read_te_typing (line 8, /tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule merging_read_te_typing (line 34, /tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule te_typing_annotation (line 51, /tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule te_typing_annotation_propagation_GRID (line 79, /tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule get_uncategorized_ins_reads_1 (line 134, /tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule get_uncategorized_ins_reads_2 (line 151, /tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule merge_uncategorized_ins_reads (line 167, /tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule index_categorized_ins_reads (line 184, /tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule locate_motif (line 197, /tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule extend_motif (line 217, /tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule extract_unc_ins_motif_subregions (line 232, /tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule rename_unc_ins_motif_subregions (line 250, /tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule concat_motif_subregions_files (line 263, /tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule clustering_unc_reads_final_set (line 276, /tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule reformat_clustering_results_1 (line 293, /tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule reformat_clustering_results_2 (line 306, /tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule final_cluster_sizes (line 323, /tmp/tmpjsaaylsc/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 0, + "subscribers_count": 1, + "topics": [], + "updated_at": 1726050889.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": null, + "formatting": "[DEBUG] \n[WARNING] In file \"/tmp/tmp2dl8hrsu/Snakefile\": Keyword \"shell\" at line 260 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp2dl8hrsu/Snakefile\": Keyword \"shell\" at line 548 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmp2dl8hrsu/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "HealthML/MFD", + "latest_release": null, + "linting": "WorkflowError in file /tmp/tmp2dl8hrsu/Snakefile, line 1:\nWorkflow defines configfile src/config.yaml but it is not present or accessible (full checked path: /tmp/tmp2dl8hrsu/src/config.yaml).\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 0, + "subscribers_count": 4, + "topics": [], + "updated_at": 1726044610.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": null, + "formatting": "[DEBUG] \n[ERROR] In file \"/tmp/tmpqoiceb5e/Snakefile\": InvalidPython: Black error:\n```\nCannot parse: 6:8: storage:\n```\n\n[INFO] In file \"/tmp/tmpqoiceb5e/Snakefile\": 1 file(s) raised parsing errors \ud83e\udd15\n\nsnakefmt version: 0.10.2", + "full_name": "dakesan/snakemake-s3-minimum-example", + "latest_release": null, + "linting": "InvalidPluginException in file /tmp/tmpqoiceb5e/Snakefile, line 6:\nError loading Snakemake plugin s3: The package snakemake-storage-plugin-s3 is not installed.\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 0, + "subscribers_count": 1, + "topics": [], + "updated_at": 1726027343.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": null, + "formatting": "[DEBUG] \n[WARNING] In file \"/tmp/tmp93c8a3i2/workflow/Snakefile\": Keyword \"output\" at line 39 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp93c8a3i2/workflow/Snakefile\": Keyword \"params\" at line 53 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmp93c8a3i2/workflow/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "MetaSUB-CAMP/camp_mag-qc", + "latest_release": null, + "linting": "KeyError in file /tmp/tmp93c8a3i2/workflow/Snakefile, line 13:\n\u0027work_dir\u0027\n File \"/tmp/tmp93c8a3i2/workflow/Snakefile\", line 13, in \u003cmodule\u003e\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 0, + "subscribers_count": 1, + "topics": [], + "updated_at": 1726024734.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": null, + "formatting": "[DEBUG] \n[WARNING] In file \"/tmp/tmpij19n6zn/workflow/Snakefile\": Keyword \"input\" at line 38 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[ERROR] In file \"/tmp/tmpij19n6zn/workflow/Snakefile\": NoParametersError: L189: In output definition.\n[INFO] In file \"/tmp/tmpij19n6zn/workflow/Snakefile\": 1 file(s) raised parsing errors \ud83e\udd15\n\nsnakefmt version: 0.10.2", + "full_name": "palomnyk/NHANES_Complexity", + "latest_release": null, + "linting": "SyntaxError in file /tmp/tmpij19n6zn/workflow/Snakefile, line 104:\nCommand must be given as string after the shell keyword.:\nNone\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 0, + "subscribers_count": 1, + "topics": [], + "updated_at": 1726014098.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": null, + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp5xd23uhl/workflow/rules/clustering.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5xd23uhl/workflow/rules/gliph.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5xd23uhl/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5xd23uhl/workflow/rules/plotting.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5xd23uhl/workflow/rules/conga.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5xd23uhl/workflow/rules/seurat_preprocessing.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5xd23uhl/workflow/rules/tcr_dist.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5xd23uhl/workflow/rules/differential_expression.smk\": Formatted content is different from original\n[INFO] 8 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "carpenter-lab/single-cell-workflow", + "latest_release": null, + "linting": "\nEidoValidationError in file /tmp/tmp5xd23uhl/workflow/Snakefile, line 13:\nEidoValidationError (Validation failed): {\"\u0027_samples\u0027 is a required property\": [{\u0027type\u0027: \"\u0027_samples\u0027 is a required property\", \u0027message\u0027: \"\u0027_samples\u0027 is a required property on instance project\", \u0027sample_name\u0027: \u0027project\u0027}]}\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/eido/validation.py\", line 73, in validate_project\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/eido/validation.py\", line 56, in _validate_object\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 0, + "subscribers_count": 1, + "topics": [], + "updated_at": 1725999338.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": null, + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpxyb6x22i/workflow/rules/phages.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpxyb6x22i/workflow/rules/binning.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpxyb6x22i/workflow/rules/coverm.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpxyb6x22i/workflow/rules/refine_bins.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpxyb6x22i/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpxyb6x22i/workflow/rules/preprocessing.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpxyb6x22i/workflow/rules/taxonomy.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmpxyb6x22i/workflow/rules/assembly.smk\": Keyword \"input\" at line 4 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmpxyb6x22i/workflow/rules/assembly.smk\": Formatted content is different from original\n[INFO] 8 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "megjohnson1999/prophage-pipeline", + "latest_release": null, + "linting": "WorkflowError in file /tmp/tmpxyb6x22i/workflow/Snakefile, line 6:\nWorkflow defines configfile ../config/config.yaml but it is not present or accessible (full checked path: /tmp/config/config.yaml).\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 0, + "subscribers_count": 1, + "topics": [], + "updated_at": 1725997832.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": null, + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp5yqp4tmy/workflow/rules/searchgui.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5yqp4tmy/workflow/rules/raw_file_parse.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5yqp4tmy/workflow/rules/generate_db.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5yqp4tmy/workflow/rules/peptide_shaker.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5yqp4tmy/workflow/rules/post_processing.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5yqp4tmy/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5yqp4tmy/workflow/rules/gff_generation.smk\": Formatted content is different from original\n[INFO] 7 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "PRIDE-reanalysis/MetaPUF", + "latest_release": null, + "linting": "NotImplementedError in file /tmp/tmp5yqp4tmy/workflow/rules/raw_file_parse.smk, line 10:\nRemote providers have been replaced by Snakemake storage plugins. Please use the corresponding storage plugin instead (snakemake-storage-plugin-*).\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 0, + "subscribers_count": 2, + "topics": [], + "updated_at": 1725982636.0 + }, + { + "config_readme": "\u003cp\u003eDescribe how to configure the workflow (using config.yaml and maybe additional files).\nAll of them need to be present with example entries inside of the config folder.\u003c/p\u003e\n", + "data_format": 2, + "description": "Generating consistent TC and TC Surge hazard set event using Snakemake and CLIMADA", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpybi507vm/workflow/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "spjuhel/Climada_TC_TCSurge_Generation", + "latest_release": null, + "linting": "KeyError in file /tmp/tmpybi507vm/workflow/Snakefile, line 27:\n\u0027start\u0027\n File \"/tmp/tmpybi507vm/workflow/Snakefile\", line 27, in \u003cmodule\u003e\n", + "mandatory_flags": { + "desc": null, + "flags": null + }, + "report": true, + "software_stack_deployment": { + "conda": true, + "singularity": false, + "singularity+conda": false + }, + "standardized": true, + "stargazers_count": 0, + "subscribers_count": 1, + "topics": [], + "updated_at": 1725981203.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": null, + "formatting": "[DEBUG] \n[DEBUG] \n[DEBUG] In file \"/tmp/tmp_8rvuhe_/workflow/rules/phylogeny.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] \n[DEBUG] \n[INFO] 1 file(s) would be changed \ud83d\ude2c\n[INFO] 4 file(s) would be left unchanged \ud83c\udf89\n\nsnakefmt version: 0.10.2", + "full_name": "koisland/CenClustering", + "latest_release": null, + "linting": "Multiple includes of /tmp/tmp_8rvuhe_/workflow/rules/data.smk ignored\nLints for snakefile https://:ghs_9K7szIWhrwDYcwcKyxn0stCzzoXvKe0s9Fdn@raw.githubusercontent.com/koisland/asm-to-reference-alignment/minimal/workflow/Snakefile:\n * Absolute path \"/tmp\" in line 13:\n Do not define absolute paths inside of the workflow, since this renders\n your workflow irreproducible on other machines. Use path relative to the\n working directory instead, or make the path configurable via a config\n file.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/configuration.html#configuration\n\nLints for snakefile /tmp/tmp_8rvuhe_/workflow/rules/phylogeny.smk:\n * Deprecated singularity directive used for container definition in line 25.:\n Use the container directive instead (it is agnostic of the underlying\n container runtime).\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule wget (line 1, /tmp/tmp_8rvuhe_/workflow/rules/utils.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule get_chm13_asm (line 6, /tmp/tmp_8rvuhe_/workflow/rules/data.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule get_chimp_asm (line 28, /tmp/tmp_8rvuhe_/workflow/rules/data.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 0, + "subscribers_count": 1, + "topics": [], + "updated_at": 1725980621.0 + }, { "config_readme": "\u003cp\u003eDescribe how to configure the workflow (using config.yaml and maybe additional files).\nAll of them need to be present with example entries inside of the config folder.\u003c/p\u003e\n", "data_format": 2, @@ -3668,26 +4032,6 @@ var data = "topics": [], "updated_at": 1725388992.0 }, - { - "config_readme": null, - "data_format": 2, - "description": "A Snakemake pipeline to analyze RNA-seq expression data", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp6w5wdo9q/workflow/rules/align.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp6w5wdo9q/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp6w5wdo9q/workflow/rules/plots.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp6w5wdo9q/workflow/rules/splicing.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp6w5wdo9q/workflow/rules/quantification.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp6w5wdo9q/workflow/rules/diffexp.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp6w5wdo9q/workflow/rules/trim.smk\": Formatted content is different from original\n[INFO] 7 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "kristinassong/RNAseq", - "latest_release": null, - "linting": "Lints for rule star_index (line 1, /tmp/tmp6w5wdo9q/workflow/rules/align.smk):\n * Param idx is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n\nLints for rule star_align (line 55, /tmp/tmp6w5wdo9q/workflow/rules/align.smk):\n * Param out_prefix is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n * Param idx is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n\nLints for rule genomecov (line 197, /tmp/tmp6w5wdo9q/workflow/rules/align.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule kallisto_quant (line 85, /tmp/tmp6w5wdo9q/workflow/rules/quantification.smk):\n * Param outdir is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n\nLints for rule tx2gene (line 142, /tmp/tmp6w5wdo9q/workflow/rules/quantification.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule merge_kallisto_quant (line 174, /tmp/tmp6w5wdo9q/workflow/rules/quantification.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule deseq2 (line 3, /tmp/tmp6w5wdo9q/workflow/rules/diffexp.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Param kallisto_dir is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n * Param out_dir is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n\nLints for rule volcano_plot (line 44, /tmp/tmp6w5wdo9q/workflow/rules/diffexp.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule pca (line 1, /tmp/tmp6w5wdo9q/workflow/rules/plots.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule multiqc (line 39, /tmp/tmp6w5wdo9q/workflow/rules/plots.smk):\n * Param outdir is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n\nLints for rule rmats (line 1, /tmp/tmp6w5wdo9q/workflow/rules/splicing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule filter_rmats (line 45, /tmp/tmp6w5wdo9q/workflow/rules/splicing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Param dir is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n\nLints for rule rmats_paired_env (line 87, /tmp/tmp6w5wdo9q/workflow/rules/splicing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule rmats_paired (line 119, /tmp/tmp6w5wdo9q/workflow/rules/splicing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule filter_rmats_paired (line 164, /tmp/tmp6w5wdo9q/workflow/rules/splicing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Param dir is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 0, - "subscribers_count": 1, - "topics": [ - "rna-seq-pipeline", - "snakemake-workflow" - ], - "updated_at": 1725385224.0 - }, { "config_readme": null, "data_format": 2, @@ -3818,23 +4162,6 @@ var data = "topics": [], "updated_at": 1725262513.0 }, - { - "config_readme": null, - "data_format": 2, - "description": "Workflow to identify the centromere dip region (CDR) from methyl BAMs", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpj9qo803b/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[INFO] 1 file(s) would be changed \ud83d\ude2c\n[INFO] 1 file(s) would be left unchanged \ud83c\udf89\n\nsnakefmt version: 0.10.2", - "full_name": "koisland/CDR-Finder", - "latest_release": null, - "linting": null, - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 0, - "subscribers_count": 1, - "topics": [], - "updated_at": 1725217590.0 - }, { "config_readme": null, "data_format": 2, @@ -4496,183 +4823,6 @@ var data = ], "updated_at": 1725457492.0 }, - { - "config_readme": null, - "data_format": 2, - "description": null, - "formatting": "[DEBUG] \n[ERROR] In file \"/tmp/tmpqoiceb5e/Snakefile\": InvalidPython: Black error:\n```\nCannot parse: 6:8: storage:\n```\n\n[INFO] In file \"/tmp/tmpqoiceb5e/Snakefile\": 1 file(s) raised parsing errors \ud83e\udd15\n\nsnakefmt version: 0.10.2", - "full_name": "dakesan/snakemake-s3-minimum-example", - "latest_release": null, - "linting": "InvalidPluginException in file /tmp/tmpqoiceb5e/Snakefile, line 6:\nError loading Snakemake plugin s3: The package snakemake-storage-plugin-s3 is not installed.\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 0, - "subscribers_count": 1, - "topics": [], - "updated_at": 1726027343.0 - }, - { - "config_readme": null, - "data_format": 2, - "description": null, - "formatting": "[DEBUG] \n[WARNING] In file \"/tmp/tmp93c8a3i2/workflow/Snakefile\": Keyword \"output\" at line 39 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp93c8a3i2/workflow/Snakefile\": Keyword \"params\" at line 53 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmp93c8a3i2/workflow/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "MetaSUB-CAMP/camp_mag-qc", - "latest_release": null, - "linting": "KeyError in file /tmp/tmp93c8a3i2/workflow/Snakefile, line 13:\n\u0027work_dir\u0027\n File \"/tmp/tmp93c8a3i2/workflow/Snakefile\", line 13, in \u003cmodule\u003e\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 0, - "subscribers_count": 1, - "topics": [], - "updated_at": 1726024734.0 - }, - { - "config_readme": null, - "data_format": 2, - "description": null, - "formatting": "[DEBUG] \n[WARNING] In file \"/tmp/tmpij19n6zn/workflow/Snakefile\": Keyword \"input\" at line 38 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[ERROR] In file \"/tmp/tmpij19n6zn/workflow/Snakefile\": NoParametersError: L189: In output definition.\n[INFO] In file \"/tmp/tmpij19n6zn/workflow/Snakefile\": 1 file(s) raised parsing errors \ud83e\udd15\n\nsnakefmt version: 0.10.2", - "full_name": "palomnyk/NHANES_Complexity", - "latest_release": null, - "linting": "SyntaxError in file /tmp/tmpij19n6zn/workflow/Snakefile, line 104:\nCommand must be given as string after the shell keyword.:\nNone\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 0, - "subscribers_count": 1, - "topics": [], - "updated_at": 1726014098.0 - }, - { - "config_readme": null, - "data_format": 2, - "description": "Opinionated viral metagenomics workflow", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpzeq5bd12/workflow/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "dhoconno/nvd", - "latest_release": null, - "linting": "KeyError in file /tmp/tmpzeq5bd12/workflow/Snakefile, line 25:\n\u0027run_id\u0027\n File \"/tmp/tmpzeq5bd12/workflow/Snakefile\", line 25, in \u003cmodule\u003e\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 0, - "subscribers_count": 1, - "topics": [], - "updated_at": 1726006482.0 - }, - { - "config_readme": null, - "data_format": 2, - "description": null, - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp5xd23uhl/workflow/rules/clustering.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5xd23uhl/workflow/rules/gliph.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5xd23uhl/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5xd23uhl/workflow/rules/plotting.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5xd23uhl/workflow/rules/conga.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5xd23uhl/workflow/rules/seurat_preprocessing.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5xd23uhl/workflow/rules/tcr_dist.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5xd23uhl/workflow/rules/differential_expression.smk\": Formatted content is different from original\n[INFO] 8 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "carpenter-lab/single-cell-workflow", - "latest_release": null, - "linting": "\nEidoValidationError in file /tmp/tmp5xd23uhl/workflow/Snakefile, line 13:\nEidoValidationError (Validation failed): {\"\u0027_samples\u0027 is a required property\": [{\u0027type\u0027: \"\u0027_samples\u0027 is a required property\", \u0027message\u0027: \"\u0027_samples\u0027 is a required property on instance project\", \u0027sample_name\u0027: \u0027project\u0027}]}\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/eido/validation.py\", line 73, in validate_project\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/eido/validation.py\", line 56, in _validate_object\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 0, - "subscribers_count": 1, - "topics": [], - "updated_at": 1725999338.0 - }, - { - "config_readme": null, - "data_format": 2, - "description": null, - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpxyb6x22i/workflow/rules/phages.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpxyb6x22i/workflow/rules/binning.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpxyb6x22i/workflow/rules/coverm.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpxyb6x22i/workflow/rules/refine_bins.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpxyb6x22i/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpxyb6x22i/workflow/rules/preprocessing.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpxyb6x22i/workflow/rules/taxonomy.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmpxyb6x22i/workflow/rules/assembly.smk\": Keyword \"input\" at line 4 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmpxyb6x22i/workflow/rules/assembly.smk\": Formatted content is different from original\n[INFO] 8 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "megjohnson1999/prophage-pipeline", - "latest_release": null, - "linting": "WorkflowError in file /tmp/tmpxyb6x22i/workflow/Snakefile, line 6:\nWorkflow defines configfile ../config/config.yaml but it is not present or accessible (full checked path: /tmp/config/config.yaml).\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 0, - "subscribers_count": 1, - "topics": [], - "updated_at": 1725997832.0 - }, - { - "config_readme": null, - "data_format": 2, - "description": null, - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp96qwng2n/workflow/rules/trim.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp96qwng2n/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp96qwng2n/workflow/rules/human_filter.smk\": Formatted content is different from original\n[DEBUG] \n[ERROR] In file \"/tmp/tmp96qwng2n/workflow/rules/test_trim.smk\": SyntaxError: L28: Unrecognised keyword \u0027ouput\u0027 in rule definition\n[INFO] In file \"/tmp/tmp96qwng2n/workflow/rules/test_trim.smk\": 1 file(s) raised parsing errors \ud83e\udd15\n[INFO] In file \"/tmp/tmp96qwng2n/workflow/rules/test_trim.smk\": 3 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "Zjx01/RNA_pipeline_Snakemake", - "latest_release": null, - "linting": "FileNotFoundError in file /tmp/tmp96qwng2n/workflow/Snakefile, line 14:\n[Errno 2] No such file or directory: \u0027/Users/zhaoj11/Documents/RNA_pipeline_Snakemake/config/config.yaml\u0027\n File \"/tmp/tmp96qwng2n/workflow/Snakefile\", line 14, in \u003cmodule\u003e\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 0, - "subscribers_count": 1, - "topics": [], - "updated_at": 1725995179.0 - }, - { - "config_readme": null, - "data_format": 2, - "description": null, - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp5yqp4tmy/workflow/rules/searchgui.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5yqp4tmy/workflow/rules/raw_file_parse.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5yqp4tmy/workflow/rules/generate_db.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5yqp4tmy/workflow/rules/peptide_shaker.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5yqp4tmy/workflow/rules/post_processing.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5yqp4tmy/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp5yqp4tmy/workflow/rules/gff_generation.smk\": Formatted content is different from original\n[INFO] 7 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "PRIDE-reanalysis/MetaPUF", - "latest_release": null, - "linting": "NotImplementedError in file /tmp/tmp5yqp4tmy/workflow/rules/raw_file_parse.smk, line 10:\nRemote providers have been replaced by Snakemake storage plugins. Please use the corresponding storage plugin instead (snakemake-storage-plugin-*).\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 0, - "subscribers_count": 2, - "topics": [], - "updated_at": 1725982636.0 - }, - { - "config_readme": "\u003cp\u003eDescribe how to configure the workflow (using config.yaml and maybe additional files).\nAll of them need to be present with example entries inside of the config folder.\u003c/p\u003e\n", - "data_format": 2, - "description": "Generating consistent TC and TC Surge hazard set event using Snakemake and CLIMADA", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpybi507vm/workflow/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "spjuhel/Climada_TC_TCSurge_Generation", - "latest_release": null, - "linting": "KeyError in file /tmp/tmpybi507vm/workflow/Snakefile, line 27:\n\u0027start\u0027\n File \"/tmp/tmpybi507vm/workflow/Snakefile\", line 27, in \u003cmodule\u003e\n", - "mandatory_flags": { - "desc": null, - "flags": null - }, - "report": true, - "software_stack_deployment": { - "conda": true, - "singularity": false, - "singularity+conda": false - }, - "standardized": true, - "stargazers_count": 0, - "subscribers_count": 1, - "topics": [], - "updated_at": 1725981203.0 - }, - { - "config_readme": null, - "data_format": 2, - "description": null, - "formatting": "[DEBUG] \n[DEBUG] \n[DEBUG] In file \"/tmp/tmp_8rvuhe_/workflow/rules/phylogeny.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] \n[DEBUG] \n[INFO] 1 file(s) would be changed \ud83d\ude2c\n[INFO] 4 file(s) would be left unchanged \ud83c\udf89\n\nsnakefmt version: 0.10.2", - "full_name": "koisland/CenClustering", - "latest_release": null, - "linting": "Multiple includes of /tmp/tmp_8rvuhe_/workflow/rules/data.smk ignored\nLints for snakefile https://:ghs_9K7szIWhrwDYcwcKyxn0stCzzoXvKe0s9Fdn@raw.githubusercontent.com/koisland/asm-to-reference-alignment/minimal/workflow/Snakefile:\n * Absolute path \"/tmp\" in line 13:\n Do not define absolute paths inside of the workflow, since this renders\n your workflow irreproducible on other machines. Use path relative to the\n working directory instead, or make the path configurable via a config\n file.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/configuration.html#configuration\n\nLints for snakefile /tmp/tmp_8rvuhe_/workflow/rules/phylogeny.smk:\n * Deprecated singularity directive used for container definition in line 25.:\n Use the container directive instead (it is agnostic of the underlying\n container runtime).\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule wget (line 1, /tmp/tmp_8rvuhe_/workflow/rules/utils.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule get_chm13_asm (line 6, /tmp/tmp_8rvuhe_/workflow/rules/data.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule get_chimp_asm (line 28, /tmp/tmp_8rvuhe_/workflow/rules/data.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 0, - "subscribers_count": 1, - "topics": [], - "updated_at": 1725980621.0 - }, { "config_readme": null, "data_format": 2, @@ -4690,30 +4840,6 @@ var data = "topics": [], "updated_at": 1725956929.0 }, - { - "config_readme": "\u003cp\u003eDescribe how to configure the workflow (using config.yaml and maybe additional files).\nAll of them need to be present with example entries inside of the config folder.\u003c/p\u003e\n", - "data_format": 2, - "description": null, - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp3zgt3_rt/workflow/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "loganylchen/snakemake-epitranscriptome", - "latest_release": null, - "linting": null, - "mandatory_flags": { - "desc": null, - "flags": null - }, - "report": true, - "software_stack_deployment": { - "conda": true, - "singularity": false, - "singularity+conda": false - }, - "standardized": true, - "stargazers_count": 0, - "subscribers_count": 1, - "topics": [], - "updated_at": 1725951944.0 - }, { "config_readme": null, "data_format": 2, @@ -4871,23 +4997,6 @@ var data = "topics": [], "updated_at": 1725890597.0 }, - { - "config_readme": null, - "data_format": 2, - "description": null, - "formatting": "[DEBUG] \n[WARNING] In file \"/tmp/tmpt6nz733o/Snakefile\": Keyword \"shell\" at line 260 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpt6nz733o/Snakefile\": Keyword \"shell\" at line 548 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmpt6nz733o/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "HealthML/MFD", - "latest_release": null, - "linting": "WorkflowError in file /tmp/tmpt6nz733o/Snakefile, line 1:\nWorkflow defines configfile src/config.yaml but it is not present or accessible (full checked path: /tmp/tmpt6nz733o/src/config.yaml).\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 0, - "subscribers_count": 4, - "topics": [], - "updated_at": 1725890300.0 - }, { "config_readme": "\u003cp\u003eDescribe how to configure the workflow (using config.yaml and maybe additional files).\nAll of them need to be present with example entries inside of the config folder.\u003c/p\u003e\n", "data_format": 2, @@ -5453,23 +5562,6 @@ var data = "topics": [], "updated_at": 1719350139.0 }, - { - "config_readme": null, - "data_format": 2, - "description": "[INSERM U981] WGS\u0026WES Pipelines", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp38ixrkfh/workflow/rules/init_pipeline.smk\": Formatted content is different from original\n[DEBUG] \n[ERROR] In file \"/tmp/tmp38ixrkfh/workflow/Snakefile\": InvalidPython: Black error:\n```\nCannot parse: 62:0: \u003cline number missing in source\u003e\n```\n\n[INFO] In file \"/tmp/tmp38ixrkfh/workflow/Snakefile\": 1 file(s) raised parsing errors \ud83e\udd15\n[INFO] In file \"/tmp/tmp38ixrkfh/workflow/Snakefile\": 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "jinxin-wang/Genome_Sequencing_Analysis", - "latest_release": null, - "linting": "Traceback (most recent call last):\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/cli.py\", line 2024, in args_to_api\n any_lint = workflow_api.lint()\n ^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/api.py\", line 337, in _handle_no_dag\n return method(self, *args, **kwargs)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/api.py\", line 354, in lint\n workflow.include(\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/workflow.py\", line 1412, in include\n exec(compile(code, snakefile.get_path_or_uri(), \"exec\"), self.globals)\n\n File \"/tmp/tmp38ixrkfh/workflow/Snakefile\", line 8, in \u003cmodule\u003e\n include: \"rules/data/utils/gz.smk\"\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/workflow.py\", line 1412, in include\n exec(compile(code, snakefile.get_path_or_uri(), \"exec\"), self.globals)\n\n File \"/tmp/tmp38ixrkfh/workflow/rules/init_pipeline.smk\", line 118, in \u003cmodule\u003e\n with open(VARIANT_CALL_TABLE,\u0027r\u0027) as fd:\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n\nFileNotFoundError: [Errno 2] No such file or directory: \u0027config/variant_call_list_TvN.tsv\u0027\n\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 0, - "subscribers_count": 2, - "topics": [], - "updated_at": 1724936205.0 - }, { "config_readme": null, "data_format": 2, @@ -6935,23 +7027,6 @@ var data = "topics": [], "updated_at": 1716626160.0 }, - { - "config_readme": null, - "data_format": 2, - "description": "Snakemake workflow to re-batch POD5 files by pore/channel", - "formatting": "[DEBUG] \n[ERROR] In file \"/tmp/tmpjl205i_1/workflow/Snakefile\": InvalidPython: Black error:\n```\nCannot parse: 68:0: res = defaultdict(set)\n```\n\n[INFO] In file \"/tmp/tmpjl205i_1/workflow/Snakefile\": 1 file(s) raised parsing errors \ud83e\udd15\n\nsnakefmt version: 0.10.2", - "full_name": "tbooth/pod5_by_pore", - "latest_release": null, - "linting": "Lints for snakefile /tmp/tmpjl205i_1/workflow/Snakefile:\n * Absolute path \"/{p5}.pod5\" in line 17:\n Do not define absolute paths inside of the workflow, since this renders\n your workflow irreproducible on other machines. Use path relative to the\n working directory instead, or make the path configurable via a config\n file.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/configuration.html#configuration\n * Absolute path \"/{foo}.pod5\" in line 41:\n Do not define absolute paths inside of the workflow, since this renders\n your workflow irreproducible on other machines. Use path relative to the\n working directory instead, or make the path configurable via a config\n file.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/configuration.html#configuration\n * Mixed rules and functions in same snakefile.:\n Small one-liner functions used only once should be defined as lambda\n expressions. Other functions should be collected in a common module, e.g.\n \u0027rules/common.smk\u0027. This makes the workflow steps more readable.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/modularization.html#includes\n * Path composition with \u0027+\u0027 in line 17:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n Also see:\n\n * Path composition with \u0027+\u0027 in line 41:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n Also see:\n\n\nLints for rule get_channels_in_pod (line 50, /tmp/tmpjl205i_1/workflow/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule batched_channels_in_pod (line 73, /tmp/tmpjl205i_1/workflow/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule get_channels_in_use (line 97, /tmp/tmpjl205i_1/workflow/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 0, - "subscribers_count": 1, - "topics": [], - "updated_at": 1716561099.0 - }, { "config_readme": null, "data_format": 2, @@ -7480,23 +7555,6 @@ var data = "topics": [], "updated_at": 1715630672.0 }, - { - "config_readme": null, - "data_format": 2, - "description": "Workflow for processing UAS imagery into data on bird location and species for near real-time monitoring in the Everglades", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpyvpe6lfd/weecology-everwatch-workflow-2e6fceb/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "weecology/everwatch-workflow", - "latest_release": "v0.2.0", - "linting": "WorkflowError:\nWorkflow defines configfile /blue/ewhite/everglades/everwatch-workflow/snakemake_config.yml but it is not present or accessible (full checked path: /blue/ewhite/everglades/everwatch-workflow/snakemake_config.yml).\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 0, - "subscribers_count": 3, - "topics": [], - "updated_at": 1715627548.0 - }, { "config_readme": null, "data_format": 2, @@ -11357,30 +11415,6 @@ var data = "topics": [], "updated_at": 1706900635.0 }, - { - "config_readme": "\u003cp\u003eDescribe how to configure the workflow (using config.yaml and maybe additional files).\nAll of them need to be present with example entries inside of the config folder.\u003c/p\u003e\n", - "data_format": 2, - "description": null, - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpjqk93gtq/workflow/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.0", - "full_name": "vkpat/short-read-mapping-and-QC-snakemake-pipeline", - "latest_release": null, - "linting": "WorkflowError:\nWorkflow defines configfile config.yaml but it is not present or accessible (full checked path: /tmp/tmpjqk93gtq/config.yaml).\n", - "mandatory_flags": { - "desc": null, - "flags": null - }, - "report": true, - "software_stack_deployment": { - "conda": true, - "singularity": false, - "singularity+conda": false - }, - "standardized": true, - "stargazers_count": 0, - "subscribers_count": 1, - "topics": [], - "updated_at": 1705604381.0 - }, { "config_readme": "\u003cdiv class=\"markdown-heading\"\u003e\u003ch1 class=\"heading-element\"\u003eGeneral configuration\u003c/h1\u003e\u003ca id=\"user-content-general-configuration\" class=\"anchor-element\" aria-label=\"Permalink: General configuration\" href=\"#general-configuration\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eSee extensive documentation on \u003ca href=\"https://proseq-etal.readthedocs.io/en/latest/\" rel=\"nofollow\"\u003ePROseq_etal\u0027s website\u003c/a\u003e for information regarding configuration.\u003c/p\u003e\n", "data_format": 2, @@ -13352,23 +13386,6 @@ var data = "topics": [], "updated_at": 1697652828.0 }, - { - "config_readme": null, - "data_format": 2, - "description": null, - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpuutsmd8f/Snakefile\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpuutsmd8f/rules/renv.smk\": Formatted content is different from original\n[INFO] 2 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.8.5", - "full_name": "deer-marketing-lab/dsms-lecture-ugc-ratings", - "latest_release": null, - "linting": "Lints for rule build_html (line 16, /tmp/tmpuutsmd8f/Snakefile):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable run_r from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable log_all from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule build_pdf (line 46, /tmp/tmpuutsmd8f/Snakefile):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable run_r from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable log_all from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule clean (line 89, /tmp/tmpuutsmd8f/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule renv_install (line 10, /tmp/tmpuutsmd8f/rules/renv.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable runR from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule renv_consent (line 22, /tmp/tmpuutsmd8f/rules/renv.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule renv_init (line 34, /tmp/tmpuutsmd8f/rules/renv.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule renv_snap (line 46, /tmp/tmpuutsmd8f/rules/renv.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule renv_restore (line 58, /tmp/tmpuutsmd8f/rules/renv.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 0, - "subscribers_count": 0, - "topics": [], - "updated_at": 1697729530.0 - }, { "config_readme": null, "data_format": 2, @@ -13752,23 +13769,6 @@ var data = "topics": [], "updated_at": 1698957156.0 }, - { - "config_readme": null, - "data_format": 2, - "description": "Mu-seq Workflow Utility ", - "formatting": "[DEBUG] \n[WARNING] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/mapping.smk\": Inline-formatted keyword \"threads\" at line 80 had its comments relocated above it.\nPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/mapping.smk\": Formatted content is different from original\n[DEBUG] \n[ERROR] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk\": InvalidPython: Black error:\n```\nCannot parse: 14:7: output:\n```\n\n[DEBUG] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk\": \n[WARNING] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/Snakefile\": Keyword \"input\" at line 93 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/Snakefile\": Keyword \"input\" at line 129 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/Snakefile\": Keyword \"input\" at line 168 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/Snakefile\": Keyword \"input\" at line 200 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[ERROR] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/ref.smk\": InvalidPython: Black error:\n```\nCannot parse: 18:4: run:\n```\n\n[DEBUG] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/ref.smk\": \n[DEBUG] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/quality_control.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/common.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/indexes_bam2sam.smk\": Formatted content is different from original\n[DEBUG] \n[ERROR] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/trimming.smk\": InvalidPython: Black error:\n```\nCannot parse: 45:4: log:\n```\n\n[DEBUG] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/trimming.smk\": \n[DEBUG] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/ref_utils.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/insertion_identification.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/duplicate_removal.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/common_utils.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/insertion_annotation.smk\": Formatted content is different from original\n[INFO] 3 file(s) raised parsing errors \ud83e\udd15\n[INFO] 10 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.8.0", - "full_name": "tgstoecker/MuWU", - "latest_release": "v1.5.0", - "linting": "Lints for rule create_read_type_check (line 4, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/trimming.smk):\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook direcive instead. Note that the script or notebook directive does\n not involve boilerplate. Similar to run, you will have direct access to\n params, input, output, and wildcards.Only use the run direcive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule prepare_insertion_table_folder_GRID (line 4, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/insertion_identification.smk):\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook direcive instead. Note that the script or notebook directive does\n not involve boilerplate. Similar to run, you will have direct access to\n params, input, output, and wildcards.Only use the run direcive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule read_te_typing (line 8, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule merging_read_te_typing (line 59, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule te_typing_annotation (line 92, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule te_typing_annotation_propagation_GRID (line 142, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule get_uncategorized_ins_reads_1 (line 238, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule get_uncategorized_ins_reads_2 (line 280, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule merge_uncategorized_ins_reads (line 320, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule index_categorized_ins_reads (line 356, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule locate_motif (line 387, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule extend_motif (line 428, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule extract_unc_ins_motif_subregions (line 459, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule rename_unc_ins_motif_subregions (line 496, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule concat_motif_subregions_files (line 527, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule clustering_unc_reads_final_set (line 558, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule reformat_clustering_results_1 (line 593, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule reformat_clustering_results_2 (line 624, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule final_cluster_sizes (line 660, /tmp/tmprzavhl6e/tgstoecker-MuWU-cbf7dbb/workflow/rules/te_typing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 0, - "subscribers_count": 1, - "topics": [], - "updated_at": 1674650130.0 - }, { "config_readme": null, "data_format": 2, @@ -36450,6 +36450,91 @@ var data = "topics": [], "updated_at": 1614368088.0 }, + { + "config_readme": null, + "data_format": 2, + "description": "CloseRead - Immunoglobulin IG sequence", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp0z3yhim4/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "phylo-lab-usc/CloseRead_ig-assembly-eval", + "latest_release": null, + "linting": "Lints for snakefile /tmp/tmp0z3yhim4/Snakefile:\n * Absolute path \"/home1/zhuyixin/zhuyixin_proj/AssmQuality\" in line 5:\n Do not define absolute paths inside of the workflow, since this renders\n your workflow irreproducible on other machines. Use path relative to the\n working directory instead, or make the path configurable via a config\n file.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/configuration.html#configuration\n * Absolute path \"/home1/zhuyixin/zhuyixin_proj/AssmQuality/igGene/{params.species}.pri.txt\" in line 76:\n Do not define absolute paths inside of the workflow, since this renders\n your workflow irreproducible on other machines. Use path relative to the\n working directory instead, or make the path configurable via a config\n file.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/configuration.html#configuration\n * Absolute path \"/home1/zhuyixin/zhuyixin_proj/AssmQuality/igGene/{params.species}.alt.txt\" in line 79:\n Do not define absolute paths inside of the workflow, since this renders\n your workflow irreproducible on other machines. Use path relative to the\n working directory instead, or make the path configurable via a config\n file.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/configuration.html#configuration\n\nLints for rule dataPrepAutomate (line 32, /tmp/tmp0z3yhim4/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule convertPrimaryBam (line 47, /tmp/tmp0z3yhim4/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule lociLocation (line 62, /tmp/tmp0z3yhim4/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule finalIGLoci (line 85, /tmp/tmp0z3yhim4/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule cigarProcessing (line 101, /tmp/tmp0z3yhim4/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule coverageAnalysis (line 126, /tmp/tmp0z3yhim4/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 1, + "subscribers_count": 1, + "topics": [], + "updated_at": 1726085317.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "A workflow for RNA sequence alignment.", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpuhscu2km/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "ryanjob42/rna-seq-align", + "latest_release": null, + "linting": "Using workflow specific profile profiles/default for setting default command line arguments.\nusage: snakemake [-h] [--dry-run] [--profile PROFILE]\n [--workflow-profile WORKFLOW_PROFILE] [--cache [RULE ...]]\n [--snakefile FILE] [--cores N] [--jobs N] [--local-cores N]\n [--resources NAME=INT [NAME=INT ...]]\n [--set-threads RULE=THREADS [RULE=THREADS ...]]\n [--max-threads MAX_THREADS]\n [--set-resources RULE:RESOURCE=VALUE [RULE:RESOURCE=VALUE ...]]\n [--set-scatter NAME=SCATTERITEMS [NAME=SCATTERITEMS ...]]\n [--set-resource-scopes RESOURCE=[global|local]\n [RESOURCE=[global|local] ...]]\n [--default-resources [NAME=INT ...]]\n [--preemptible-rules [PREEMPTIBLE_RULES ...]]\n [--preemptible-retries PREEMPTIBLE_RETRIES]\n [--configfile FILE [FILE ...]] [--config [KEY=VALUE ...]]\n [--envvars VARNAME [VARNAME ...]] [--directory DIR] [--touch]\n [--keep-going]\n [--rerun-triggers {code,input,mtime,params,software-env} [{code,input,mtime,params,software-env} ...]]\n [--force] [--executor {local,dryrun,touch}] [--forceall]\n [--forcerun [TARGET ...]] [--prioritize TARGET [TARGET ...]]\n [--batch RULE=BATCH/BATCHES] [--until TARGET [TARGET ...]]\n [--omit-from TARGET [TARGET ...]] [--rerun-incomplete]\n [--shadow-prefix DIR] [--scheduler [{ilp,greedy}]]\n [--wms-monitor [WMS_MONITOR]]\n [--wms-monitor-arg [NAME=VALUE ...]]\n [--scheduler-ilp-solver {PULP_CBC_CMD,COIN_CMD}]\n [--conda-base-path CONDA_BASE_PATH] [--no-subworkflows]\n [--precommand PRECOMMAND] [--groups GROUPS [GROUPS ...]]\n [--group-components GROUP_COMPONENTS [GROUP_COMPONENTS ...]]\n [--report [FILE]] [--report-stylesheet CSSFILE]\n [--reporter PLUGIN] [--draft-notebook TARGET]\n [--edit-notebook TARGET] [--notebook-listen IP:PORT]\n [--lint [{text,json}]] [--generate-unit-tests [TESTPATH]]\n [--containerize] [--export-cwl FILE] [--list-rules]\n [--list-target-rules] [--dag] [--rulegraph] [--filegraph]\n [--d3dag] [--summary] [--detailed-summary] [--archive FILE]\n [--cleanup-metadata FILE [FILE ...]] [--cleanup-shadow]\n [--skip-script-cleanup] [--unlock]\n [--list-changes {input,params,code}] [--list-input-changes]\n [--list-params-changes] [--list-untracked]\n [--delete-all-output | --delete-temp-output]\n [--keep-incomplete] [--drop-metadata] [--version]\n [--printshellcmds] [--debug-dag] [--nocolor]\n [--quiet [{all,host,progress,rules} ...]]\n [--print-compilation] [--verbose] [--force-use-threads]\n [--allow-ambiguity] [--nolock] [--ignore-incomplete]\n [--max-inventory-time SECONDS] [--latency-wait SECONDS]\n [--wait-for-files [FILE ...]] [--wait-for-files-file FILE]\n [--queue-input-wait-time SECONDS] [--notemp] [--all-temp]\n [--unneeded-temp-files FILE [FILE ...]]\n [--keep-storage-local-copies]\n [--target-files-omit-workdir-adjustment]\n [--allowed-rules ALLOWED_RULES [ALLOWED_RULES ...]]\n [--max-jobs-per-timespan MAX_JOBS_PER_TIMESPAN]\n [--max-jobs-per-second MAX_JOBS_PER_SECOND]\n [--max-status-checks-per-second MAX_STATUS_CHECKS_PER_SECOND]\n [--seconds-between-status-checks SECONDS_BETWEEN_STATUS_CHECKS]\n [--retries RETRIES] [--wrapper-prefix WRAPPER_PREFIX]\n [--default-storage-provider DEFAULT_STORAGE_PROVIDER]\n [--default-storage-prefix DEFAULT_STORAGE_PREFIX]\n [--local-storage-prefix LOCAL_STORAGE_PREFIX]\n [--remote-job-local-storage-prefix REMOTE_JOB_LOCAL_STORAGE_PREFIX]\n [--shared-fs-usage {input-output,persistence,software-deployment,source-cache,sources,storage-local-copies,none} [{input-output,persistence,software-deployment,source-cache,sources,storage-local-copies,none} ...]]\n [--scheduler-greediness SCHEDULER_GREEDINESS] [--no-hooks]\n [--debug] [--runtime-profile FILE]\n [--local-groupid LOCAL_GROUPID] [--attempt ATTEMPT]\n [--show-failed-logs] [--log-handler-script FILE]\n [--log-service {none,slack,wms}] [--job-deploy-sources]\n [--benchmark-extended] [--container-image IMAGE]\n [--immediate-submit] [--jobscript SCRIPT] [--jobname NAME]\n [--flux]\n [--software-deployment-method {apptainer,conda,env-modules} [{apptainer,conda,env-modules} ...]]\n [--container-cleanup-images] [--use-conda]\n [--conda-not-block-search-path-envvars] [--list-conda-envs]\n [--conda-prefix DIR] [--conda-cleanup-envs]\n [--conda-cleanup-pkgs [{tarballs,cache}]]\n [--conda-create-envs-only] [--conda-frontend {conda,mamba}]\n [--use-apptainer] [--apptainer-prefix DIR]\n [--apptainer-args ARGS] [--use-envmodules]\n [--scheduler-solver-path SCHEDULER_SOLVER_PATH]\n [--deploy-sources QUERY CHECKSUM]\n [--target-jobs TARGET_JOBS [TARGET_JOBS ...]]\n [--mode {subprocess,default,remote}]\n [--report-html-path VALUE]\n [--report-html-stylesheet-path VALUE]\n [targets ...]\nsnakemake: error: argument --executor/-e: invalid choice: \u0027slurm\u0027 (choose from \u0027local\u0027, \u0027dryrun\u0027, \u0027touch\u0027)\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 1, + "subscribers_count": 2, + "topics": [], + "updated_at": 1726082338.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "NIRISS Pure Parallel Processing Scripts", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp61ywvmuw/workflow/rules/contamination.smk\": Formatted content is different from original\n[DEBUG] \n[ERROR] In file \"/tmp/tmp61ywvmuw/workflow/rules/download.smk\": InvalidPython: Black error:\n```\nCannot parse: 3:0: \u003cline number missing in source\u003e\n```\n\n[DEBUG] In file \"/tmp/tmp61ywvmuw/workflow/rules/download.smk\": \n[DEBUG] In file \"/tmp/tmp61ywvmuw/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp61ywvmuw/workflow/rules/extract.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp61ywvmuw/workflow/rules/preprocess.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp61ywvmuw/workflow/rules/mosaic.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp61ywvmuw/workflow/rules/makeFitsmap.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp61ywvmuw/workflow/rules/redshiftFit.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp61ywvmuw/workflow/rules/runStage1.smk\": Formatted content is different from original\n[INFO] 1 file(s) raised parsing errors \ud83e\udd15\n[INFO] 8 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "OutThere-JWST/NISPureParallel", + "latest_release": null, + "linting": "Using workflow specific profile profiles/default for setting default command line arguments.\nModuleNotFoundError in file /tmp/tmp61ywvmuw/workflow/Snakefile, line 6:\nNo module named \u0027astropy\u0027\n File \"/tmp/tmp61ywvmuw/workflow/Snakefile\", line 6, in \u003cmodule\u003e\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 1, + "subscribers_count": 3, + "topics": [], + "updated_at": 1726079621.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "Opinionated viral metagenomics workflow", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp_s8amaf5/workflow/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "dhoconno/nvd", + "latest_release": null, + "linting": "KeyError in file /tmp/tmp_s8amaf5/workflow/Snakefile, line 31:\n\u0027run_id\u0027\n File \"/tmp/tmp_s8amaf5/workflow/Snakefile\", line 31, in \u003cmodule\u003e\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 1, + "subscribers_count": 1, + "topics": [], + "updated_at": 1726077730.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "Collection of rules used to call structural variants", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp082hw9pr/hydra-genetics-cnv_sv-0467058/workflow/rules/cnvpytor.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp082hw9pr/hydra-genetics-cnv_sv-0467058/workflow/rules/exomedepth.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp082hw9pr/hydra-genetics-cnv_sv-0467058/workflow/rules/tiddit.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp082hw9pr/hydra-genetics-cnv_sv-0467058/workflow/rules/reviewer.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp082hw9pr/hydra-genetics-cnv_sv-0467058/workflow/rules/purecn.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp082hw9pr/hydra-genetics-cnv_sv-0467058/workflow/rules/expansionhunter.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp082hw9pr/hydra-genetics-cnv_sv-0467058/workflow/rules/pindel.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp082hw9pr/hydra-genetics-cnv_sv-0467058/workflow/rules/smncopynumbercaller.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp082hw9pr/hydra-genetics-cnv_sv-0467058/workflow/rules/cnvkit.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp082hw9pr/hydra-genetics-cnv_sv-0467058/workflow/rules/upd.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp082hw9pr/hydra-genetics-cnv_sv-0467058/workflow/rules/common.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp082hw9pr/hydra-genetics-cnv_sv-0467058/workflow/rules/manta.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp082hw9pr/hydra-genetics-cnv_sv-0467058/workflow/rules/gatk.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp082hw9pr/hydra-genetics-cnv_sv-0467058/workflow/rules/svdb.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] \n[DEBUG] In file \"/tmp/tmp082hw9pr/hydra-genetics-cnv_sv-0467058/workflow/rules/automap.smk\": Formatted content is different from original\n[INFO] 15 file(s) would be changed \ud83d\ude2c\n[INFO] 1 file(s) would be left unchanged \ud83c\udf89\n\nsnakefmt version: 0.10.2", + "full_name": "hydra-genetics/cnv_sv", + "latest_release": "v0.5.0", + "linting": "ModuleNotFoundError in file /tmp/tmp082hw9pr/hydra-genetics-cnv_sv-0467058/workflow/rules/common.smk, line 13:\nNo module named \u0027hydra_genetics\u0027\n File \"/tmp/tmp082hw9pr/hydra-genetics-cnv_sv-0467058/workflow/rules/common.smk\", line 13, in \u003cmodule\u003e\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 1, + "subscribers_count": 4, + "topics": [], + "updated_at": 1726034384.0 + }, { "config_readme": null, "data_format": 2, @@ -37515,23 +37600,6 @@ var data = "topics": [], "updated_at": 1725032821.0 }, - { - "config_readme": null, - "data_format": 2, - "description": "A workflow for RNA sequence alignment.", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpq7d3si8m/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "ryanjob42/rna-seq-align", - "latest_release": null, - "linting": "Using workflow specific profile profiles/default for setting default command line arguments.\nusage: snakemake [-h] [--dry-run] [--profile PROFILE]\n [--workflow-profile WORKFLOW_PROFILE] [--cache [RULE ...]]\n [--snakefile FILE] [--cores N] [--jobs N] [--local-cores N]\n [--resources NAME=INT [NAME=INT ...]]\n [--set-threads RULE=THREADS [RULE=THREADS ...]]\n [--max-threads MAX_THREADS]\n [--set-resources RULE:RESOURCE=VALUE [RULE:RESOURCE=VALUE ...]]\n [--set-scatter NAME=SCATTERITEMS [NAME=SCATTERITEMS ...]]\n [--set-resource-scopes RESOURCE=[global|local]\n [RESOURCE=[global|local] ...]]\n [--default-resources [NAME=INT ...]]\n [--preemptible-rules [PREEMPTIBLE_RULES ...]]\n [--preemptible-retries PREEMPTIBLE_RETRIES]\n [--configfile FILE [FILE ...]] [--config [KEY=VALUE ...]]\n [--envvars VARNAME [VARNAME ...]] [--directory DIR] [--touch]\n [--keep-going]\n [--rerun-triggers {code,input,mtime,params,software-env} [{code,input,mtime,params,software-env} ...]]\n [--force] [--executor {local,dryrun,touch}] [--forceall]\n [--forcerun [TARGET ...]] [--prioritize TARGET [TARGET ...]]\n [--batch RULE=BATCH/BATCHES] [--until TARGET [TARGET ...]]\n [--omit-from TARGET [TARGET ...]] [--rerun-incomplete]\n [--shadow-prefix DIR] [--scheduler [{ilp,greedy}]]\n [--wms-monitor [WMS_MONITOR]]\n [--wms-monitor-arg [NAME=VALUE ...]]\n [--scheduler-ilp-solver {PULP_CBC_CMD,COIN_CMD}]\n [--conda-base-path CONDA_BASE_PATH] [--no-subworkflows]\n [--precommand PRECOMMAND] [--groups GROUPS [GROUPS ...]]\n [--group-components GROUP_COMPONENTS [GROUP_COMPONENTS ...]]\n [--report [FILE]] [--report-stylesheet CSSFILE]\n [--reporter PLUGIN] [--draft-notebook TARGET]\n [--edit-notebook TARGET] [--notebook-listen IP:PORT]\n [--lint [{text,json}]] [--generate-unit-tests [TESTPATH]]\n [--containerize] [--export-cwl FILE] [--list-rules]\n [--list-target-rules] [--dag] [--rulegraph] [--filegraph]\n [--d3dag] [--summary] [--detailed-summary] [--archive FILE]\n [--cleanup-metadata FILE [FILE ...]] [--cleanup-shadow]\n [--skip-script-cleanup] [--unlock]\n [--list-changes {params,code,input}] [--list-input-changes]\n [--list-params-changes] [--list-untracked]\n [--delete-all-output | --delete-temp-output]\n [--keep-incomplete] [--drop-metadata] [--version]\n [--printshellcmds] [--debug-dag] [--nocolor]\n [--quiet [{all,host,progress,rules} ...]]\n [--print-compilation] [--verbose] [--force-use-threads]\n [--allow-ambiguity] [--nolock] [--ignore-incomplete]\n [--max-inventory-time SECONDS] [--latency-wait SECONDS]\n [--wait-for-files [FILE ...]] [--wait-for-files-file FILE]\n [--queue-input-wait-time SECONDS] [--notemp] [--all-temp]\n [--unneeded-temp-files FILE [FILE ...]]\n [--keep-storage-local-copies]\n [--target-files-omit-workdir-adjustment]\n [--allowed-rules ALLOWED_RULES [ALLOWED_RULES ...]]\n [--max-jobs-per-timespan MAX_JOBS_PER_TIMESPAN]\n [--max-jobs-per-second MAX_JOBS_PER_SECOND]\n [--max-status-checks-per-second MAX_STATUS_CHECKS_PER_SECOND]\n [--seconds-between-status-checks SECONDS_BETWEEN_STATUS_CHECKS]\n [--retries RETRIES] [--wrapper-prefix WRAPPER_PREFIX]\n [--default-storage-provider DEFAULT_STORAGE_PROVIDER]\n [--default-storage-prefix DEFAULT_STORAGE_PREFIX]\n [--local-storage-prefix LOCAL_STORAGE_PREFIX]\n [--remote-job-local-storage-prefix REMOTE_JOB_LOCAL_STORAGE_PREFIX]\n [--shared-fs-usage {input-output,persistence,software-deployment,source-cache,sources,storage-local-copies,none} [{input-output,persistence,software-deployment,source-cache,sources,storage-local-copies,none} ...]]\n [--scheduler-greediness SCHEDULER_GREEDINESS] [--no-hooks]\n [--debug] [--runtime-profile FILE]\n [--local-groupid LOCAL_GROUPID] [--attempt ATTEMPT]\n [--show-failed-logs] [--log-handler-script FILE]\n [--log-service {none,slack,wms}] [--job-deploy-sources]\n [--benchmark-extended] [--container-image IMAGE]\n [--immediate-submit] [--jobscript SCRIPT] [--jobname NAME]\n [--flux]\n [--software-deployment-method {apptainer,conda,env-modules} [{apptainer,conda,env-modules} ...]]\n [--container-cleanup-images] [--use-conda]\n [--conda-not-block-search-path-envvars] [--list-conda-envs]\n [--conda-prefix DIR] [--conda-cleanup-envs]\n [--conda-cleanup-pkgs [{tarballs,cache}]]\n [--conda-create-envs-only] [--conda-frontend {conda,mamba}]\n [--use-apptainer] [--apptainer-prefix DIR]\n [--apptainer-args ARGS] [--use-envmodules]\n [--scheduler-solver-path SCHEDULER_SOLVER_PATH]\n [--deploy-sources QUERY CHECKSUM]\n [--target-jobs TARGET_JOBS [TARGET_JOBS ...]]\n [--mode {default,subprocess,remote}]\n [--report-html-path VALUE]\n [--report-html-stylesheet-path VALUE]\n [targets ...]\nsnakemake: error: argument --executor/-e: invalid choice: \u0027slurm\u0027 (choose from \u0027local\u0027, \u0027dryrun\u0027, \u0027touch\u0027)\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 1, - "subscribers_count": 2, - "topics": [], - "updated_at": 1726001552.0 - }, { "config_readme": null, "data_format": 2, @@ -37566,40 +37634,6 @@ var data = "topics": [], "updated_at": 1725745131.0 }, - { - "config_readme": null, - "data_format": 2, - "description": "CloseRead - Immunoglobulin IG sequence", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpgu_iycgc/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "phylo-lab-usc/CloseRead_ig-assembly-eval", - "latest_release": null, - "linting": "Lints for snakefile /tmp/tmpgu_iycgc/Snakefile:\n * Absolute path \"/home1/zhuyixin/zhuyixin_proj/AssmQuality\" in line 5:\n Do not define absolute paths inside of the workflow, since this renders\n your workflow irreproducible on other machines. Use path relative to the\n working directory instead, or make the path configurable via a config\n file.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/configuration.html#configuration\n * Absolute path \"/home1/zhuyixin/zhuyixin_proj/AssmQuality/igGene/{params.species}.pri.txt\" in line 76:\n Do not define absolute paths inside of the workflow, since this renders\n your workflow irreproducible on other machines. Use path relative to the\n working directory instead, or make the path configurable via a config\n file.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/configuration.html#configuration\n * Absolute path \"/home1/zhuyixin/zhuyixin_proj/AssmQuality/igGene/{params.species}.alt.txt\" in line 79:\n Do not define absolute paths inside of the workflow, since this renders\n your workflow irreproducible on other machines. Use path relative to the\n working directory instead, or make the path configurable via a config\n file.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/configuration.html#configuration\n\nLints for rule dataPrepAutomate (line 32, /tmp/tmpgu_iycgc/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule convertPrimaryBam (line 47, /tmp/tmpgu_iycgc/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule lociLocation (line 62, /tmp/tmpgu_iycgc/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule finalIGLoci (line 85, /tmp/tmpgu_iycgc/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule cigarProcessing (line 101, /tmp/tmpgu_iycgc/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule coverageAnalysis (line 126, /tmp/tmpgu_iycgc/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable HOME from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 1, - "subscribers_count": 1, - "topics": [], - "updated_at": 1726024286.0 - }, - { - "config_readme": null, - "data_format": 2, - "description": "NIRISS Pure Parallel Processing Scripts", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpww5a1o8q/workflow/rules/extract.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpww5a1o8q/workflow/rules/contamination.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpww5a1o8q/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpww5a1o8q/workflow/rules/runStage1.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpww5a1o8q/workflow/rules/makeFitsmap.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpww5a1o8q/workflow/rules/mosaic.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpww5a1o8q/workflow/rules/preprocess.smk\": Formatted content is different from original\n[DEBUG] \n[ERROR] In file \"/tmp/tmpww5a1o8q/workflow/rules/download.smk\": InvalidPython: Black error:\n```\nCannot parse: 3:0: \u003cline number missing in source\u003e\n```\n\n[DEBUG] In file \"/tmp/tmpww5a1o8q/workflow/rules/download.smk\": \n[DEBUG] In file \"/tmp/tmpww5a1o8q/workflow/rules/redshiftFit.smk\": Formatted content is different from original\n[INFO] 1 file(s) raised parsing errors \ud83e\udd15\n[INFO] 8 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "OutThere-JWST/NISPureParallel", - "latest_release": null, - "linting": "Using workflow specific profile profiles/default for setting default command line arguments.\nModuleNotFoundError in file /tmp/tmpww5a1o8q/workflow/Snakefile, line 5:\nNo module named \u0027astropy\u0027\n File \"/tmp/tmpww5a1o8q/workflow/Snakefile\", line 5, in \u003cmodule\u003e\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 1, - "subscribers_count": 3, - "topics": [], - "updated_at": 1726001584.0 - }, { "config_readme": null, "data_format": 2, @@ -37617,23 +37651,6 @@ var data = "topics": [], "updated_at": 1725973311.0 }, - { - "config_readme": null, - "data_format": 2, - "description": "Collection of rules used to call structural variants", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp9da2tugs/hydra-genetics-cnv_sv-0467058/workflow/rules/gatk.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp9da2tugs/hydra-genetics-cnv_sv-0467058/workflow/rules/smncopynumbercaller.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp9da2tugs/hydra-genetics-cnv_sv-0467058/workflow/rules/svdb.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp9da2tugs/hydra-genetics-cnv_sv-0467058/workflow/rules/cnvkit.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp9da2tugs/hydra-genetics-cnv_sv-0467058/workflow/rules/cnvpytor.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp9da2tugs/hydra-genetics-cnv_sv-0467058/workflow/rules/upd.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp9da2tugs/hydra-genetics-cnv_sv-0467058/workflow/rules/purecn.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp9da2tugs/hydra-genetics-cnv_sv-0467058/workflow/rules/pindel.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp9da2tugs/hydra-genetics-cnv_sv-0467058/workflow/rules/common.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp9da2tugs/hydra-genetics-cnv_sv-0467058/workflow/rules/reviewer.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp9da2tugs/hydra-genetics-cnv_sv-0467058/workflow/rules/tiddit.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp9da2tugs/hydra-genetics-cnv_sv-0467058/workflow/rules/exomedepth.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp9da2tugs/hydra-genetics-cnv_sv-0467058/workflow/rules/automap.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp9da2tugs/hydra-genetics-cnv_sv-0467058/workflow/rules/manta.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp9da2tugs/hydra-genetics-cnv_sv-0467058/workflow/rules/expansionhunter.smk\": Formatted content is different from original\n[DEBUG] \n[INFO] 15 file(s) would be changed \ud83d\ude2c\n[INFO] 1 file(s) would be left unchanged \ud83c\udf89\n\nsnakefmt version: 0.10.2", - "full_name": "hydra-genetics/cnv_sv", - "latest_release": "v0.5.0", - "linting": "Traceback (most recent call last):\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/cli.py\", line 2024, in args_to_api\n any_lint = workflow_api.lint()\n ^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/api.py\", line 337, in _handle_no_dag\n return method(self, *args, **kwargs)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/api.py\", line 354, in lint\n workflow.include(\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/workflow.py\", line 1412, in include\n exec(compile(code, snakefile.get_path_or_uri(), \"exec\"), self.globals)\n\n File \"/tmp/tmp9da2tugs/hydra-genetics-cnv_sv-0467058/workflow/Snakefile\", line 7, in \u003cmodule\u003e\n include: \"rules/common.smk\"\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/workflow.py\", line 1412, in include\n exec(compile(code, snakefile.get_path_or_uri(), \"exec\"), self.globals)\n\n File \"/tmp/tmp9da2tugs/hydra-genetics-cnv_sv-0467058/workflow/rules/common.smk\", line 13, in \u003cmodule\u003e\n from hydra_genetics.utils.resources import load_resources\n\nModuleNotFoundError: No module named \u0027hydra_genetics\u0027\n\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 1, - "subscribers_count": 4, - "topics": [], - "updated_at": 1725641574.0 - }, { "config_readme": null, "data_format": 2, @@ -43214,6 +43231,104 @@ var data = "topics": [], "updated_at": 1551028733.0 }, + { + "config_readme": null, + "data_format": 2, + "description": "Estimaci\u00f3n de turbidez en el agua a la entrada de la planta de tratamiento de SAMEEP, utilizando los productos Sentinel-2 MSI L2A y aprendizaje autom\u00e1tico.", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpeac2b74k/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "vhgauto/sameep", + "latest_release": null, + "linting": "Lints for rule descarga_y_extraccion (line 8, /tmp/tmpeac2b74k/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule render_index (line 21, /tmp/tmpeac2b74k/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 2, + "subscribers_count": 1, + "topics": [ + "github-actions", + "modeling", + "python", + "r", + "random-forest", + "sentinel-2", + "snakemake", + "turbidity", + "water", + "water-quality", + "calidad-agua", + "gistaq", + "s2-msi", + "turbidez" + ], + "updated_at": 1726055066.0 + }, + { + "config_readme": "\u003cp\u003eAdd datasets to datasets.tsv, and update config.yml to customize options.\u003c/p\u003e\n", + "data_format": 2, + "description": "A Snakemake workflow for pre-processing single plane illumination microscopy (SPIM, aka lightsheet microscopy)", + "formatting": null, + "full_name": "khanlab/SPIMprep", + "latest_release": "v0.1.0-alpha", + "linting": "ModuleNotFoundError in file /tmp/tmpe99k7uer/khanlab-SPIMprep-85f518d/workflow/Snakefile, line 2:\nNo module named \u0027upath\u0027\n File \"/tmp/tmpe99k7uer/khanlab-SPIMprep-85f518d/workflow/Snakefile\", line 2, in \u003cmodule\u003e\n", + "mandatory_flags": { + "desc": null, + "flags": null + }, + "report": true, + "software_stack_deployment": { + "conda": false, + "singularity": true, + "singularity+conda": false + }, + "standardized": true, + "stargazers_count": 2, + "subscribers_count": 2, + "topics": [], + "updated_at": 1725992911.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "Circrnas in Host And viRuses anaLysis pIpEline for Detection Annotation Quantification of circRNAs", + "formatting": "[DEBUG] \n[WARNING] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/findcircrna.smk\": Keyword \"params\" at line 200 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/findcircrna.smk\": Keyword \"params\" at line 377 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\;\u0027\n[DEBUG] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/findcircrna.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/init.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/align.smk\": Keyword \"output\" at line 38 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/align.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/Snakefile\": Keyword \"input\" at line 293 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/post_findcircrna_processing.smk\": Keyword \"input\" at line 356 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/post_findcircrna_processing.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/create_index.smk\": Formatted content is different from original\n[DEBUG] \n[INFO] 6 file(s) would be changed \ud83d\ude2c\n[INFO] 3 file(s) would be left unchanged \ud83c\udf89\n\nsnakefmt version: 0.10.2", + "full_name": "CCBR/CHARLIE", + "latest_release": "v0.10.1", + "linting": "/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/init.smk:141: SyntaxWarning: invalid escape sequence \u0027\\/\u0027\n d = d.strip(\"r\\/\")\nKeyError in file /tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/init.smk, line 92:\n\u0027workdir\u0027\n File \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/init.smk\", line 92, in \u003cmodule\u003e\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 2, + "subscribers_count": 3, + "topics": [ + "circrna", + "circularrna-detection", + "ngs-analysis", + "ngs-pipeline" + ], + "updated_at": 1725991438.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "Caprion pilot study", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp70oohp0q/workflow/rules/cojo.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmp70oohp0q/workflow/rules/prs.smk\": Keyword \"input\" at line 28 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmp70oohp0q/workflow/rules/prs.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp70oohp0q/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp70oohp0q/workflow/rules/test.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmp70oohp0q/workflow/rules/rg.smk\": Keyword \"input\" at line 33 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmp70oohp0q/workflow/rules/rg.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp70oohp0q/workflow/rules/report.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmp70oohp0q/workflow/rules/mr.smk\": Keyword \"input\" at line 43 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[ERROR] In file \"/tmp/tmp70oohp0q/workflow/rules/mr.smk\": InvalidParameterSyntax: 97ref_col=lambda wil dcards: OUTCOMES.loc[wildcards.OutcomeCode][\u0027COLUMNS\u0027][\u0027REF\u0027]\n\n[INFO] In file \"/tmp/tmp70oohp0q/workflow/rules/mr.smk\": 1 file(s) raised parsing errors \ud83e\udd15\n[INFO] In file \"/tmp/tmp70oohp0q/workflow/rules/mr.smk\": 6 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "jinghuazhao/Caprion", + "latest_release": null, + "linting": "Traceback (most recent call last):\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/cli.py\", line 2003, in args_to_api\n any_lint = workflow_api.lint()\n ^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/api.py\", line 291, in _handle_no_dag\n return method(self, *args, **kwargs)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/api.py\", line 354, in lint\n workflow.include(\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/workflow.py\", line 1435, in include\n code, rulecount = parse(\n ^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/parser.py\", line 1310, in parse\n for t, orig_token in automaton.consume():\n ^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/parser.py\", line 144, in consume\n for t, orig in self.state(token):\n ^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/parser.py\", line 1252, in python\n for t in self.subautomaton(token.string, token=token).consume():\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/parser.py\", line 171, in subautomaton\n self.error(\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/parser.py\", line 164, in error\n raise SyntaxError(msg, (self.snakefile.path, lineno(token), None, None))\n\n File \"/tmp/tmp70oohp0q/workflow/Snakefile\", line 8\n\nSyntaxError: Keyword subworkflow is deprecated. Use module directive instead (see docs).\n\n\nDuring handling of the above exception, another exception occurred:\n\n\nTraceback (most recent call last):\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/cli.py\", line 2173, in main\n success = args_to_api(args, parser)\n ^^^^^^^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/cli.py\", line 2147, in args_to_api\n snakemake_api.print_exception(e)\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/api.py\", line 239, in print_exception\n print_exception(ex, linemaps)\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/exceptions.py\", line 96, in print_exception\n format_error(\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/exceptions.py\", line 20, in format_error\n lineno = linemaps[snakefile][lineno]\n ~~~~~~~~~~~~~~~~~~~^^^^^^^^\n\nKeyError: 8\n\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 2, + "subscribers_count": 3, + "topics": [ + "proteomics", + "gwas" + ], + "updated_at": 1725983365.0 + }, { "config_readme": null, "data_format": 2, @@ -43895,38 +44010,6 @@ var data = ], "updated_at": 1725813958.0 }, - { - "config_readme": null, - "data_format": 2, - "description": "Estimaci\u00f3n de turbidez en el agua a la entrada de la planta de tratamiento de SAMEEP, utilizando los productos Sentinel-2 MSI L2A y aprendizaje autom\u00e1tico.", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpbj2dwyx4/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "vhgauto/sameep", - "latest_release": null, - "linting": "Lints for rule descarga_y_extraccion (line 15, /tmp/tmpbj2dwyx4/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule render_index (line 46, /tmp/tmpbj2dwyx4/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 2, - "subscribers_count": 1, - "topics": [ - "github-actions", - "modeling", - "python", - "r", - "random-forest", - "sentinel-2", - "snakemake", - "turbidity", - "water", - "water-quality", - "calidad-agua", - "gistaq", - "s2-msi", - "turbidez" - ], - "updated_at": 1725617034.0 - }, { "config_readme": null, "data_format": 2, @@ -43980,72 +44063,6 @@ var data = "topics": [], "updated_at": 1725538673.0 }, - { - "config_readme": "\u003cp\u003eAdd datasets to datasets.tsv, and update config.yml to customize options.\u003c/p\u003e\n", - "data_format": 2, - "description": "A Snakemake workflow for pre-processing single plane illumination microscopy (SPIM, aka lightsheet microscopy)", - "formatting": null, - "full_name": "khanlab/SPIMprep", - "latest_release": "v0.1.0-alpha", - "linting": "ModuleNotFoundError in file /tmp/tmpe99k7uer/khanlab-SPIMprep-85f518d/workflow/Snakefile, line 2:\nNo module named \u0027upath\u0027\n File \"/tmp/tmpe99k7uer/khanlab-SPIMprep-85f518d/workflow/Snakefile\", line 2, in \u003cmodule\u003e\n", - "mandatory_flags": { - "desc": null, - "flags": null - }, - "report": true, - "software_stack_deployment": { - "conda": false, - "singularity": true, - "singularity+conda": false - }, - "standardized": true, - "stargazers_count": 2, - "subscribers_count": 2, - "topics": [], - "updated_at": 1725992911.0 - }, - { - "config_readme": null, - "data_format": 2, - "description": "Circrnas in Host And viRuses anaLysis pIpEline for Detection Annotation Quantification of circRNAs", - "formatting": "[DEBUG] \n[WARNING] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/findcircrna.smk\": Keyword \"params\" at line 200 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/findcircrna.smk\": Keyword \"params\" at line 377 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\;\u0027\n[DEBUG] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/findcircrna.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/init.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/align.smk\": Keyword \"output\" at line 38 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/align.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/Snakefile\": Keyword \"input\" at line 293 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/post_findcircrna_processing.smk\": Keyword \"input\" at line 356 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/post_findcircrna_processing.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/create_index.smk\": Formatted content is different from original\n[DEBUG] \n[INFO] 6 file(s) would be changed \ud83d\ude2c\n[INFO] 3 file(s) would be left unchanged \ud83c\udf89\n\nsnakefmt version: 0.10.2", - "full_name": "CCBR/CHARLIE", - "latest_release": "v0.10.1", - "linting": "/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/init.smk:141: SyntaxWarning: invalid escape sequence \u0027\\/\u0027\n d = d.strip(\"r\\/\")\nKeyError in file /tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/init.smk, line 92:\n\u0027workdir\u0027\n File \"/tmp/tmp38klwqhg/CCBR-CHARLIE-38aae11/workflow/rules/init.smk\", line 92, in \u003cmodule\u003e\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 2, - "subscribers_count": 3, - "topics": [ - "circrna", - "circularrna-detection", - "ngs-analysis", - "ngs-pipeline" - ], - "updated_at": 1725991438.0 - }, - { - "config_readme": null, - "data_format": 2, - "description": "Caprion pilot study", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp70oohp0q/workflow/rules/cojo.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmp70oohp0q/workflow/rules/prs.smk\": Keyword \"input\" at line 28 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmp70oohp0q/workflow/rules/prs.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp70oohp0q/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp70oohp0q/workflow/rules/test.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmp70oohp0q/workflow/rules/rg.smk\": Keyword \"input\" at line 33 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmp70oohp0q/workflow/rules/rg.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp70oohp0q/workflow/rules/report.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmp70oohp0q/workflow/rules/mr.smk\": Keyword \"input\" at line 43 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[ERROR] In file \"/tmp/tmp70oohp0q/workflow/rules/mr.smk\": InvalidParameterSyntax: 97ref_col=lambda wil dcards: OUTCOMES.loc[wildcards.OutcomeCode][\u0027COLUMNS\u0027][\u0027REF\u0027]\n\n[INFO] In file \"/tmp/tmp70oohp0q/workflow/rules/mr.smk\": 1 file(s) raised parsing errors \ud83e\udd15\n[INFO] In file \"/tmp/tmp70oohp0q/workflow/rules/mr.smk\": 6 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "jinghuazhao/Caprion", - "latest_release": null, - "linting": "Traceback (most recent call last):\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/cli.py\", line 2003, in args_to_api\n any_lint = workflow_api.lint()\n ^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/api.py\", line 291, in _handle_no_dag\n return method(self, *args, **kwargs)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/api.py\", line 354, in lint\n workflow.include(\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/workflow.py\", line 1435, in include\n code, rulecount = parse(\n ^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/parser.py\", line 1310, in parse\n for t, orig_token in automaton.consume():\n ^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/parser.py\", line 144, in consume\n for t, orig in self.state(token):\n ^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/parser.py\", line 1252, in python\n for t in self.subautomaton(token.string, token=token).consume():\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/parser.py\", line 171, in subautomaton\n self.error(\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/parser.py\", line 164, in error\n raise SyntaxError(msg, (self.snakefile.path, lineno(token), None, None))\n\n File \"/tmp/tmp70oohp0q/workflow/Snakefile\", line 8\n\nSyntaxError: Keyword subworkflow is deprecated. Use module directive instead (see docs).\n\n\nDuring handling of the above exception, another exception occurred:\n\n\nTraceback (most recent call last):\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/cli.py\", line 2173, in main\n success = args_to_api(args, parser)\n ^^^^^^^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/cli.py\", line 2147, in args_to_api\n snakemake_api.print_exception(e)\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/api.py\", line 239, in print_exception\n print_exception(ex, linemaps)\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/exceptions.py\", line 96, in print_exception\n format_error(\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/exceptions.py\", line 20, in format_error\n lineno = linemaps[snakefile][lineno]\n ~~~~~~~~~~~~~~~~~~~^^^^^^^^\n\nKeyError: 8\n\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 2, - "subscribers_count": 3, - "topics": [ - "proteomics", - "gwas" - ], - "updated_at": 1725983365.0 - }, { "config_readme": null, "data_format": 2, @@ -46767,6 +46784,57 @@ var data = ], "updated_at": 1617804426.0 }, + { + "config_readme": null, + "data_format": 2, + "description": "A model of the Illinois electricity system built with PyPSA.", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpbse_9wfv/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "ucsusa/pypsa-illinois", + "latest_release": null, + "linting": "ModuleNotFoundError in file /tmp/tmpbse_9wfv/Snakefile, line 5:\nNo module named \u0027dotenv\u0027\n File \"/tmp/tmpbse_9wfv/Snakefile\", line 5, in \u003cmodule\u003e\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 3, + "subscribers_count": 1, + "topics": [], + "updated_at": 1726083747.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "Variant calling workflow for mixed ploidy samples implemented in Snakemake and using GATK4 ", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpxs3nt0j_/workflow/rules/0_indexreference.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmpxs3nt0j_/workflow/rules/4_filter_bcftools.smk\": Keyword \"output\" at line 98 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmpxs3nt0j_/workflow/rules/4_filter_bcftools.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpxs3nt0j_/workflow/rules/4_filter_gatk.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmpxs3nt0j_/workflow/rules/1_mapreads.smk\": Keyword \"input\" at line 417 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmpxs3nt0j_/workflow/rules/1_mapreads.smk\": Formatted content is different from original\n[DEBUG] \n[ERROR] In file \"/tmp/tmpxs3nt0j_/workflow/rules/3_genotypeGVCF.smk\": SyntaxError: L130: Unrecognised keyword \u0027vcf_stats_GT_DP_multiallelic\u0027 in rule definition\n[DEBUG] In file \"/tmp/tmpxs3nt0j_/workflow/rules/3_genotypeGVCF.smk\": \n[DEBUG] In file \"/tmp/tmpxs3nt0j_/workflow/rules/2_callvars.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpxs3nt0j_/workflow/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) raised parsing errors \ud83e\udd15\n[INFO] 6 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "jgerchen/polyploid_variant_calling", + "latest_release": null, + "linting": "KeyError in file /tmp/tmpxs3nt0j_/workflow/rules/0_indexreference.smk, line 6:\n\u0027input_fasta\u0027\n File \"/tmp/tmpxs3nt0j_/workflow/rules/0_indexreference.smk\", line 6, in \u003cmodule\u003e\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 3, + "subscribers_count": 1, + "topics": [], + "updated_at": 1726043720.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "This repository contains the source code for the datasets used in the implementation of the dengue lineage system in NextClade. The organizational workflow was developed based on the workflow created for the m-pox virus, available at: https://github.com/nextstrain/mpox/tree/master/nextclade.\"", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpb5lq7c6v/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "V-GEN-Lab/dengue-lineages-workflow", + "latest_release": null, + "linting": "Lints for snakefile /tmp/tmpb5lq7c6v/Snakefile:\n * Mixed rules and functions in same snakefile.:\n Small one-liner functions used only once should be defined as lambda\n expressions. Other functions should be collected in a common module, e.g.\n \u0027rules/common.smk\u0027. This makes the workflow steps more readable.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/modularization.html#includes\n\nLints for rule augur_ancestral (line 27, /tmp/tmpb5lq7c6v/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule augur_translate (line 47, /tmp/tmpb5lq7c6v/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule augur_clades (line 64, /tmp/tmpb5lq7c6v/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule colors (line 83, /tmp/tmpb5lq7c6v/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule augur_export (line 99, /tmp/tmpb5lq7c6v/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule assemble_dataset (line 123, /tmp/tmpb5lq7c6v/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule clean (line 151, /tmp/tmpb5lq7c6v/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 3, + "subscribers_count": 0, + "topics": [], + "updated_at": 1725978690.0 + }, { "config_readme": null, "data_format": 2, @@ -47143,57 +47211,6 @@ var data = "topics": [], "updated_at": 1725479591.0 }, - { - "config_readme": null, - "data_format": 2, - "description": "A model of the Illinois electricity system built with PyPSA.", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpgtdshvzm/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "ucsusa/pypsa-illinois", - "latest_release": null, - "linting": "Traceback (most recent call last):\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/cli.py\", line 2024, in args_to_api\n any_lint = workflow_api.lint()\n ^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/api.py\", line 337, in _handle_no_dag\n return method(self, *args, **kwargs)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/api.py\", line 354, in lint\n workflow.include(\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/workflow.py\", line 1412, in include\n exec(compile(code, snakefile.get_path_or_uri(), \"exec\"), self.globals)\n\n File \"/tmp/tmpgtdshvzm/Snakefile\", line 7, in \u003cmodule\u003e\n\nModuleNotFoundError: No module named \u0027dotenv\u0027\n\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 3, - "subscribers_count": 1, - "topics": [], - "updated_at": 1725459211.0 - }, - { - "config_readme": null, - "data_format": 2, - "description": "This repository contains the source code for the datasets used in the implementation of the dengue lineage system in NextClade. The organizational workflow was developed based on the workflow created for the m-pox virus, available at: https://github.com/nextstrain/mpox/tree/master/nextclade.\"", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpb5lq7c6v/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "V-GEN-Lab/dengue-lineages-workflow", - "latest_release": null, - "linting": "Lints for snakefile /tmp/tmpb5lq7c6v/Snakefile:\n * Mixed rules and functions in same snakefile.:\n Small one-liner functions used only once should be defined as lambda\n expressions. Other functions should be collected in a common module, e.g.\n \u0027rules/common.smk\u0027. This makes the workflow steps more readable.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/modularization.html#includes\n\nLints for rule augur_ancestral (line 27, /tmp/tmpb5lq7c6v/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule augur_translate (line 47, /tmp/tmpb5lq7c6v/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule augur_clades (line 64, /tmp/tmpb5lq7c6v/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule colors (line 83, /tmp/tmpb5lq7c6v/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule augur_export (line 99, /tmp/tmpb5lq7c6v/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule assemble_dataset (line 123, /tmp/tmpb5lq7c6v/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule clean (line 151, /tmp/tmpb5lq7c6v/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 3, - "subscribers_count": 0, - "topics": [], - "updated_at": 1725978690.0 - }, - { - "config_readme": null, - "data_format": 2, - "description": "Variant calling workflow for mixed ploidy samples implemented in Snakemake and using GATK4 ", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpyfp0t2_1/workflow/rules/4_filter_gatk.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpyfp0t2_1/workflow/Snakefile\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpyfp0t2_1/workflow/rules/2_callvars.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpyfp0t2_1/workflow/rules/0_indexreference.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmpyfp0t2_1/workflow/rules/4_filter_bcftools.smk\": Keyword \"output\" at line 98 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmpyfp0t2_1/workflow/rules/4_filter_bcftools.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmpyfp0t2_1/workflow/rules/1_mapreads.smk\": Keyword \"input\" at line 417 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmpyfp0t2_1/workflow/rules/1_mapreads.smk\": Formatted content is different from original\n[DEBUG] \n[ERROR] In file \"/tmp/tmpyfp0t2_1/workflow/rules/3_genotypeGVCF.smk\": SyntaxError: L130: Unrecognised keyword \u0027vcf_stats_GT_DP_multiallelic\u0027 in rule definition\n[INFO] In file \"/tmp/tmpyfp0t2_1/workflow/rules/3_genotypeGVCF.smk\": 1 file(s) raised parsing errors \ud83e\udd15\n[INFO] In file \"/tmp/tmpyfp0t2_1/workflow/rules/3_genotypeGVCF.smk\": 6 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "jgerchen/polyploid_variant_calling", - "latest_release": null, - "linting": "KeyError in file /tmp/tmpyfp0t2_1/workflow/rules/0_indexreference.smk, line 6:\n\u0027input_fasta\u0027\n File \"/tmp/tmpyfp0t2_1/workflow/rules/0_indexreference.smk\", line 6, in \u003cmodule\u003e\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 3, - "subscribers_count": 1, - "topics": [], - "updated_at": 1725975931.0 - }, { "config_readme": null, "data_format": 2, @@ -48801,6 +48818,57 @@ var data = "topics": [], "updated_at": 1616871931.0 }, + { + "config_readme": null, + "data_format": 2, + "description": "Snakemake workflow for the preprocessing, alignment, QC, and quantification of spatial transcriptomics data", + "formatting": "[DEBUG] \n[WARNING] In file \"/tmp/tmpnbreyozq/rules/6a_scanpy_init.smk\": Keyword \"input\" at line 8 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpnbreyozq/rules/6a_scanpy_init.smk\": Keyword \"params\" at line 12 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpnbreyozq/rules/6a_scanpy_init.smk\": Keyword \"input\" at line 126 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] \n[DEBUG] \n[ERROR] In file \"/tmp/tmpnbreyozq/Snakefile\": InvalidPython: Black error:\n```\nCannot parse: 249:0: EOF in multi-line statement\n```\n\n[DEBUG] In file \"/tmp/tmpnbreyozq/Snakefile\": \n[DEBUG] In file \"/tmp/tmpnbreyozq/rules/0_utils.smk\": Formatted content is different from original\n[INFO] 1 file(s) raised parsing errors \ud83e\udd15\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n[INFO] 2 file(s) would be left unchanged \ud83c\udf89\n\nsnakefmt version: 0.10.2", + "full_name": "mckellardw/slide_snake", + "latest_release": null, + "linting": "Using workflow specific profile profiles/default for setting default command line arguments.\nLints for snakefile /tmp/tmpnbreyozq/rules/0_utils.smk:\n * Mixed rules and functions in same snakefile.:\n Small one-liner functions used only once should be defined as lambda\n expressions. Other functions should be collected in a common module, e.g.\n \u0027rules/common.smk\u0027. This makes the workflow steps more readable.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/modularization.html#includes\n\nLints for snakefile /tmp/tmpnbreyozq/rules/0a_barcode_maps.smk:\n * Path composition with \u0027+\u0027 in line 7:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 75:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n\nLints for rule utils_index_BAM (line 5, /tmp/tmpnbreyozq/rules/0_utils.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule BC_copy_barcode_map (line 5, /tmp/tmpnbreyozq/rules/0a_barcode_maps.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule BC_get_simple_whitelist (line 29, /tmp/tmpnbreyozq/rules/0a_barcode_maps.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule BC_write_whitelist_variants (line 47, /tmp/tmpnbreyozq/rules/0a_barcode_maps.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule BC_insert_adapter_into_list (line 139, /tmp/tmpnbreyozq/rules/0a_barcode_maps.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule ilmn_1a_merge_fastqs (line 2, /tmp/tmpnbreyozq/rules/short_read/1a_mergefqs.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule ilmn_1b_R1_hardTrimming (line 136, /tmp/tmpnbreyozq/rules/short_read/1b_trimming.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ilmn_1b_R1_internalTrimming (line 170, /tmp/tmpnbreyozq/rules/short_read/1b_trimming.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ilmn_1c_fastQC_preTrim (line 5, /tmp/tmpnbreyozq/rules/short_read/1c_fastqc.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule ilmn_1c_fastQC_postTrim (line 30, /tmp/tmpnbreyozq/rules/short_read/1c_fastqc.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule ilmn_1c_fastQC_twiceTrim (line 54, /tmp/tmpnbreyozq/rules/short_read/1c_fastqc.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule ilmn_2a_bwa_rRNA_filter_R1 (line 60, /tmp/tmpnbreyozq/rules/short_read/2a_rRNA_bwa.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ilmn_2a_bwa_rRNA_compress_unmapped (line 82, /tmp/tmpnbreyozq/rules/short_read/2a_rRNA_bwa.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ilmn_2a_bwa_rRNA_filtered_fastqc (line 99, /tmp/tmpnbreyozq/rules/short_read/2a_rRNA_bwa.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule ilmn_2b_STAR_rRNA_align (line 7, /tmp/tmpnbreyozq/rules/short_read/2b_rRNA_STAR.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule ilmn_2b_STAR_rRNA_compress_outs (line 63, /tmp/tmpnbreyozq/rules/short_read/2b_rRNA_STAR.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n * Param GENEDIR is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n\nLints for rule ilmn_2b_STAR_rRNA_rename_compress_unmapped (line 99, /tmp/tmpnbreyozq/rules/short_read/2b_rRNA_STAR.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule ilmn_2b_STAR_rRNA_filtered_fastqc (line 122, /tmp/tmpnbreyozq/rules/short_read/2b_rRNA_STAR.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule ilmn_2c_qualimapQC_rRNA_STAR (line 37, /tmp/tmpnbreyozq/rules/short_read/2c_rRNA_qualimap.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule ilmn_2c_qualimap_summary2csv_rRNA_STAR (line 66, /tmp/tmpnbreyozq/rules/short_read/2c_rRNA_qualimap.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ilmn_2d_ribodetector_get_noRibo_list (line 40, /tmp/tmpnbreyozq/rules/short_read/2d_ribodetector.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ilmn_2d_ribodetector_gunzip_R1 (line 54, /tmp/tmpnbreyozq/rules/short_read/2d_ribodetector.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ilmn_2d_ribodetector_filter_R1 (line 66, /tmp/tmpnbreyozq/rules/short_read/2d_ribodetector.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ilmn_2d_ribodetector_filter_R1_internalTrim (line 82, /tmp/tmpnbreyozq/rules/short_read/2d_ribodetector.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ilmn_2d_ribodetector_filter_R1_hardTrim (line 98, /tmp/tmpnbreyozq/rules/short_read/2d_ribodetector.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ilmn_2d_ribodetector_compress_fqs (line 114, /tmp/tmpnbreyozq/rules/short_read/2d_ribodetector.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ilmn_3a_STARsolo_align (line 9, /tmp/tmpnbreyozq/rules/short_read/3a_star_align.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule ilmn_3a_compress_STAR_outs (line 68, /tmp/tmpnbreyozq/rules/short_read/3a_star_align.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n * Param VELDIR is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n * Param GENEDIR is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n * Param GENEFULLDIR is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n\nLints for rule ilmn_3b_fastqc_unmapped (line 7, /tmp/tmpnbreyozq/rules/short_read/3b_star_unmapped.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule ilmn_3c_strand_split_dedup_bam (line 30, /tmp/tmpnbreyozq/rules/short_read/3c_star_dedup.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ilmn_3d_qualimapQC_STAR (line 3, /tmp/tmpnbreyozq/rules/short_read/3d_star_qualimap.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule ilmn_3d_qualimap_summary2csv_STAR (line 31, /tmp/tmpnbreyozq/rules/short_read/3d_star_qualimap.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ilmn_4a_kbpython_std_remove_suffix (line 65, /tmp/tmpnbreyozq/rules/short_read/4a_kbpython.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule kbpython_std_compress_outs (line 81, /tmp/tmpnbreyozq/rules/short_read/4a_kbpython.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_1a_merge_formats (line 5, /tmp/tmpnbreyozq/rules/ont/1a_preprocessing.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_1a_readIDs_by_adapter_type (line 83, /tmp/tmpnbreyozq/rules/ont/1a_preprocessing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_1a_adapter_scan_results (line 107, /tmp/tmpnbreyozq/rules/ont/1a_preprocessing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_1a_merge_scan_lists (line 129, /tmp/tmpnbreyozq/rules/ont/1a_preprocessing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_1a_subset_fastq_by_adapter_type (line 146, /tmp/tmpnbreyozq/rules/ont/1a_preprocessing.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_1a_compress_merged_fq (line 175, /tmp/tmpnbreyozq/rules/ont/1a_preprocessing.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_1b_R1_hardTrimming (line 78, /tmp/tmpnbreyozq/rules/ont/1b_trimming.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_1b_R1_internalTrim (line 108, /tmp/tmpnbreyozq/rules/ont/1b_trimming.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_1c_fastq_call_bc_from_adapter (line 3, /tmp/tmpnbreyozq/rules/ont/1c_barcode_calling.smk):\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule ont_1c_filter_read_barcodes (line 64, /tmp/tmpnbreyozq/rules/ont/1c_barcode_calling.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_1c_tsv_bc_correction (line 79, /tmp/tmpnbreyozq/rules/ont/1c_barcode_calling.smk):\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule ont_1d_sort_compress_output (line 44, /tmp/tmpnbreyozq/rules/ont/1d_minimap2.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_1d_add_featureCounts_to_bam (line 136, /tmp/tmpnbreyozq/rules/ont/1d_minimap2.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_1d_add_corrected_barcodes (line 165, /tmp/tmpnbreyozq/rules/ont/1d_minimap2.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_1d_add_umis (line 193, /tmp/tmpnbreyozq/rules/ont/1d_minimap2.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_1d_filter_bam_empty_tags (line 221, /tmp/tmpnbreyozq/rules/ont/1d_minimap2.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_1d_counts_to_sparse (line 277, /tmp/tmpnbreyozq/rules/ont/1d_minimap2.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule ont_clipBeforeSTAR (line 4, /tmp/tmpnbreyozq/rules/ont/1d_STAR.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule ont_STARsolo_align (line 27, /tmp/tmpnbreyozq/rules/ont/1d_STAR.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule ont_compress_STAR_outs (line 119, /tmp/tmpnbreyozq/rules/ont/1d_STAR.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n * Param VELDIR is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n * Param GENEFULLDIR is a prefix of input or output file but hardcoded:\n If this is meant to represent a file path prefix, it will fail when\n running workflow in environments without a shared filesystem. Instead,\n provide a function that infers the appropriate prefix from the input or\n output file, e.g.: lambda w, input: os.path.splitext(input[0])[0]\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n https://snakemake.readthedocs.io/en/stable/tutorial/advanced.html#tutorial-input-functions\n\nLints for rule ont_2_qualimap_minimap2 (line 2, /tmp/tmpnbreyozq/rules/ont/2_qualimap.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule ont_2_qualimap_STAR (line 31, /tmp/tmpnbreyozq/rules/ont/2_qualimap.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule ont_2_qualimap_summary2csv (line 60, /tmp/tmpnbreyozq/rules/ont/2_qualimap.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_readQC_0_rawInput (line 2, /tmp/tmpnbreyozq/rules/ont/2_read_qc.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_readQC_1_preCutadapt (line 26, /tmp/tmpnbreyozq/rules/ont/2_read_qc.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_readQC_2_postCutadapt (line 50, /tmp/tmpnbreyozq/rules/ont/2_read_qc.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_readQC_3_bam (line 89, /tmp/tmpnbreyozq/rules/ont/2_read_qc.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule readQC_downsample (line 116, /tmp/tmpnbreyozq/rules/ont/2_read_qc.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule ont_readQC_summaryplot (line 133, /tmp/tmpnbreyozq/rules/ont/2_read_qc.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule ont_fastQC_preTrim (line 2, /tmp/tmpnbreyozq/rules/ont/2_fastqc.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule ont_fastQC_preCutadapt (line 25, /tmp/tmpnbreyozq/rules/ont/2_fastqc.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule ont_fastQC_postCutadapt (line 48, /tmp/tmpnbreyozq/rules/ont/2_fastqc.smk):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 4, + "subscribers_count": 3, + "topics": [], + "updated_at": 1726069701.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "Knowledge Graph generator for WorkflowHub", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpf31gp8nu/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "workflowhub-eu/workflowhub-graph", + "latest_release": null, + "linting": "Lints for rule source_ro_crates (line 12, /tmp/tmpf31gp8nu/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable OUTPUT_DIRS from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable VERSIONS from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable OUTPUT_DIRS from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable VERSIONS from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable OUTPUT_DIRS from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule report_created_files (line 35, /tmp/tmpf31gp8nu/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule merge_files (line 44, /tmp/tmpf31gp8nu/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule create_ro_crate (line 70, /tmp/tmpf31gp8nu/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 4, + "subscribers_count": 9, + "topics": [], + "updated_at": 1726052509.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "Knowledge Graph generator for WorkflowHub", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmph7kt2vg4/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "workflowhub-eu/workflowhub-graph", + "latest_release": null, + "linting": "Lints for rule source_ro_crates (line 12, /tmp/tmph7kt2vg4/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable OUTPUT_DIRS from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable VERSIONS from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable OUTPUT_DIRS from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable VERSIONS from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable OUTPUT_DIRS from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule report_created_files (line 35, /tmp/tmph7kt2vg4/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule merge_files (line 44, /tmp/tmph7kt2vg4/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule create_ro_crate (line 70, /tmp/tmph7kt2vg4/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 4, + "subscribers_count": 9, + "topics": [], + "updated_at": 1726052509.0 + }, { "config_readme": "\u003cdiv class=\"markdown-heading\"\u003e\u003ch1 class=\"heading-element\"\u003eGeneral settings\u003c/h1\u003e\u003ca id=\"user-content-general-settings\" class=\"anchor\" aria-label=\"Permalink: General settings\" href=\"#general-settings\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eTo configure this workflow, modify \u003ccode\u003econfig/config.yaml\u003c/code\u003e according to your needs, following the explanations provided in the file.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch1 class=\"heading-element\"\u003eSample sheet\u003c/h1\u003e\u003ca id=\"user-content-sample-sheet\" class=\"anchor\" aria-label=\"Permalink: Sample sheet\" href=\"#sample-sheet\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eAdd samples to \u003ccode\u003econfig/samples.tsv\u003c/code\u003e. For each sample, the columns \u003ccode\u003esample_name\u003c/code\u003e, \u003ccode\u003ealias\u003c/code\u003e, \u003ccode\u003eplatform\u003c/code\u003e, and \u003ccode\u003egroup\u003c/code\u003e have to be defined.\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eSamples within the same \u003ccode\u003egroup\u003c/code\u003e will be called jointly.\u003c/li\u003e\n\u003cli\u003eAliases represent the name of the sample within its group (they can be the same as the sample name, or something simpler, e.g. tumor or normal).\u003c/li\u003e\n\u003cli\u003eThe \u003ccode\u003eplatform\u003c/code\u003e column needs to contain the used sequencing plaform (one of \u0027CAPILLARY\u0027, \u0027LS454\u0027, \u0027ILLUMINA\u0027, \u0027SOLID\u0027, \u0027HELICOS\u0027, \u0027IONTORRENT\u0027, \u0027ONT\u0027, \u0027PACBIO\u2019).\u003c/li\u003e\n\u003cli\u003eThe \u003ccode\u003effpe\u003c/code\u003e column specifies whether a sample is a ffpe substrate (1) or not (0). ffpe treated normal samples are not supported.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eMissing values can be specified by empty columns or by writing \u003ccode\u003eNA\u003c/code\u003e. Lines can be commented out with \u003ccode\u003e#\u003c/code\u003e.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch1 class=\"heading-element\"\u003eUnit sheet\u003c/h1\u003e\u003ca id=\"user-content-unit-sheet\" class=\"anchor\" aria-label=\"Permalink: Unit sheet\" href=\"#unit-sheet\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eFor each sample, add one or more sequencing units (runs, lanes or replicates) to the unit sheet \u003ccode\u003econfig/units.tsv\u003c/code\u003e.\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eEach unit has a \u003ccode\u003eunit_name\u003c/code\u003e, which can be e.g. a running number, or an actual run, lane or replicate id.\u003c/li\u003e\n\u003cli\u003eEach unit has a \u003ccode\u003esample_name\u003c/code\u003e, which associates it with the biological sample it comes from.\u003c/li\u003e\n\u003cli\u003eFor each unit, define either one (column \u003ccode\u003efq1\u003c/code\u003e) or two (columns \u003ccode\u003efq1\u003c/code\u003e, \u003ccode\u003efq2\u003c/code\u003e) FASTQ files (these can point to anywhere in your system).\u003c/li\u003e\n\u003cli\u003eAlternatively, you can define an SRA (sequence read archive) accession (starting with e.g. ERR or SRR) by using a column \u003ccode\u003esra\u003c/code\u003e. In the latter case, the pipeline will automatically download the corresponding paired end reads from SRA. If both local files and SRA accession are available, the local files will be preferred.\u003c/li\u003e\n\u003cli\u003eDefine adapters in the \u003ccode\u003eadapters\u003c/code\u003e column, by putting \u003ca href=\"https://cutadapt.readthedocs.org\" rel=\"nofollow\"\u003ecutadapt arguments\u003c/a\u003e in quotation marks (e.g. \u003ccode\u003e\"-a ACGCGATCG -A GCTAGCGTACT\"\u003c/code\u003e).\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eMissing values can be specified by empty columns or by writing \u003ccode\u003eNA\u003c/code\u003e. Lines can be commented out with \u003ccode\u003e#\u003c/code\u003e.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch1 class=\"heading-element\"\u003ePrimer trimming\u003c/h1\u003e\u003ca id=\"user-content-primer-trimming\" class=\"anchor\" aria-label=\"Permalink: Primer trimming\" href=\"#primer-trimming\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eFor panel data the pipeline allows trimming of amplicon primers on both ends of a fragment but also on a single end only.\nIn case of single end primers these are supposed to be located at the left end of a read.\nWhen primer trimming is enabled, primers have to be defined either directly in the \u003ccode\u003econfig.yaml\u003c/code\u003e or in a seperate tsv-file.\nDefining primers directly in the config file is prefered when all samples come from the same primer set.\nIn case of different panels, primers have to be set panel-wise in a seperate tsv-file (the path to that tsv can be set in the config under \u003ccode\u003eprimers/trimming/tsv\u003c/code\u003e).\nFor each panel the following columns need to be set: \u003ccode\u003epanel\u003c/code\u003e, \u003ccode\u003efa1\u003c/code\u003e and \u003ccode\u003efa2\u003c/code\u003e (optional).\nAdditionally, for each sample the corresponding panel must be defined in \u003ccode\u003esamples.tsv\u003c/code\u003e (column \u003ccode\u003epanel\u003c/code\u003e).\nFor single primer trimming only, the first entry in the config (respective in the tsv file) needs to be defined.\u003c/p\u003e\n", "data_format": 2, @@ -48998,23 +49066,6 @@ var data = "topics": [], "updated_at": 1725609533.0 }, - { - "config_readme": null, - "data_format": 2, - "description": "Snakemake workflow for the preprocessing, alignment, QC, and quantification of spatial transcriptomics data", - "formatting": "[DEBUG] \n[DEBUG] \n[ERROR] In file \"/tmp/tmpvjehfjty/Snakefile\": InvalidPython: Black error:\n```\nCannot parse: 249:0: EOF in multi-line statement\n```\n\n[DEBUG] In file \"/tmp/tmpvjehfjty/Snakefile\": \n[DEBUG] In file \"/tmp/tmpvjehfjty/rules/0_utils.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmpvjehfjty/rules/6a_scanpy_init.smk\": Keyword \"input\" at line 8 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpvjehfjty/rules/6a_scanpy_init.smk\": Keyword \"params\" at line 12 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpvjehfjty/rules/6a_scanpy_init.smk\": Keyword \"input\" at line 126 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[INFO] 1 file(s) raised parsing errors \ud83e\udd15\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n[INFO] 2 file(s) would be left unchanged \ud83c\udf89\n\nsnakefmt version: 0.10.2", - "full_name": "mckellardw/slide_snake", - "latest_release": null, - "linting": "Using workflow specific profile profiles/default for setting default command line arguments.\nKeyError in file /tmp/tmpvjehfjty/rules/ont/1d_minimap2.smk, line 51:\n\u0027REF_GENOME_FASTA\u0027\n File \"/tmp/tmpvjehfjty/rules/ont/1d_minimap2.smk\", line 51, in \u003cmodule\u003e\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 4, - "subscribers_count": 3, - "topics": [], - "updated_at": 1726002285.0 - }, { "config_readme": "\u003cdiv class=\"markdown-heading\"\u003e\u003ch1 class=\"heading-element\"\u003eGeneral Settings\u003c/h1\u003e\u003ca id=\"user-content-general-settings\" class=\"anchor\" aria-label=\"Permalink: General Settings\" href=\"#general-settings\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eIn the \u003ca href=\"./\"\u003e\u003ccode\u003econfig\u003c/code\u003e\u003c/a\u003e directory, you will find template configuration files\nfor this pipeline. For your run, you\u0027ll need to edit \u003ccode\u003esamples.tsv\u003c/code\u003e,\n\u003ccode\u003eunits.tsv\u003c/code\u003e, and \u003ccode\u003econfig.yaml\u003c/code\u003e.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch2 class=\"heading-element\"\u003eSamples list\u003c/h2\u003e\u003ca id=\"user-content-samples-list\" class=\"anchor\" aria-label=\"Permalink: Samples list\" href=\"#samples-list\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eAll your samples should be listed in \u003ccode\u003esamples.tsv\u003c/code\u003e. Samples preceded with a \u003ccode\u003e#\u003c/code\u003e\nwill not be included, this can be useful if you want to exclude a sample after\nquality checking.\u003c/p\u003e\n\u003cp\u003eEach sample must have four columns filled in the sample sheet. The columns are\ntab separated:\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003esample\u003c/code\u003e - The name of the sample\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003epopulation\u003c/code\u003e - The populations you will group your samples into. These are\nthe groups that population level stats are calculated on.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003etime\u003c/code\u003e - This should be either \u003ccode\u003emodern\u003c/code\u003e or \u003ccode\u003ehistorical\u003c/code\u003e, the only thing this\nwill affect is whether or not your bam files will be corrected for\npost-mortem damage or not.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003edepth\u003c/code\u003e - This is only used for depth filtering. Extreme levels (both high\nand low) will be calculated for each group you put here as well as the\ndataset as a whole, and all will be filtered out for all analyses. Any string\ncan be used for this. If all samples are sequenced to roughly similar depths,\nall can have the same value. If some are low coverage, and some higher,\nsimply using \u0027low\u0027 and \u0027high\u0027 on the corresponding samples is sufficient.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eThe values in the sample list will end up in filenames, so ensure that they are\nonly characters permitted in filenames on your system.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch2 class=\"heading-element\"\u003eUnits list\u003c/h2\u003e\u003ca id=\"user-content-units-list\" class=\"anchor\" aria-label=\"Permalink: Units list\" href=\"#units-list\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eAll your raw data will be pointed to in \u003ccode\u003eunits.tsv\u003c/code\u003e.\u003c/p\u003e\n\u003cp\u003eEach sample must have five columns filled in the units sheet. The columns are\ntab separated:\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003esample\u003c/code\u003e - The sample name, same as in \u003ccode\u003esamples.tsv\u003c/code\u003e.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eunit\u003c/code\u003e - This is used to fill out the \u003ccode\u003eID\u003c/code\u003e read group in the bam file. It\nmust be unique to each read group, so the same sample shouldn\u0027t have the\nsame unit for more than one sequencing run. A good format might be\n\u003ccode\u003esequencer_barcode.lane\u003c/code\u003e. Optical duplicates will be removed within units.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003elib\u003c/code\u003e - This is used to fill out the \u003ccode\u003eLB\u003c/code\u003e read group. This should be a unique\nidentifier for each sample library. Sequencing runs from the same library,\nbut different runs, will have the same value in \u003ccode\u003elib\u003c/code\u003e, but different in\n\u003ccode\u003eunit\u003c/code\u003e.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eplatform\u003c/code\u003e - This is used to fill out the \u003ccode\u003ePL\u003c/code\u003e read group. Put what you\u0027d\nwant there. Usually \u003ccode\u003eILLUMINA\u003c/code\u003e for Illumina platforms.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003efq1\u003c/code\u003e and \u003ccode\u003efq2\u003c/code\u003e - The absolute or relative paths from the working directory\nto the raw fastq files for the sample. Currently the pipeline only supports\npaired-end sequencing, single end may be added down the line.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch2 class=\"heading-element\"\u003eConfiguration file\u003c/h2\u003e\u003ca id=\"user-content-configuration-file\" class=\"anchor\" aria-label=\"Permalink: Configuration file\" href=\"#configuration-file\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003e\u003ccode\u003econfig.yaml\u003c/code\u003e contains the configuration for the workflow, this is where you\nwill put what analyses, filters, and options you want. Below I describe the\nconfiguration options. The \u003ca href=\"config.yaml\"\u003e\u003ccode\u003econfig.yaml\u003c/code\u003e\u003c/a\u003e in this repository\nserves as a template, but includes some \u0027default\u0027 parameters that may be good\nstarting points for some users. If \u003ccode\u003e--configfile\u003c/code\u003e is not specified in the\nsnakemake command, the workflow will default to \u003ccode\u003econfig/config.yaml\u003c/code\u003e.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch3 class=\"heading-element\"\u003eConfiguration options\u003c/h3\u003e\u003ca id=\"user-content-configuration-options\" class=\"anchor\" aria-label=\"Permalink: Configuration options\" href=\"#configuration-options\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch4 class=\"heading-element\"\u003eDataset Configuration\u003c/h4\u003e\u003ca id=\"user-content-dataset-configuration\" class=\"anchor\" aria-label=\"Permalink: Dataset Configuration\" href=\"#dataset-configuration\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eRequired configuration of the \u0027dataset\u0027.\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003esamples:\u003c/code\u003e An absolute or relative path from the working directory to the\n\u003ccode\u003esamples.tsv\u003c/code\u003e file.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eunits:\u003c/code\u003e An absolute or relative path from the working directory to the\n\u003ccode\u003eunits.tsv\u003c/code\u003e file.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003edataset:\u003c/code\u003e A name for this dataset run - essentially, an identifier for a\nbatch of samples to be analysed together with the same configuration.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eHere, dataset means a set of samples and configurations that the workflow will\nbe run with. Each dataset should have its own \u003ccode\u003esamples.tsv\u003c/code\u003e and \u003ccode\u003econfig.yaml\u003c/code\u003e,\nbut the same \u003ccode\u003eunits.tsv\u003c/code\u003e can be used for multiple if you prefer. Essentially,\nwhat the dataset identifier does is keeps your outputs organized into projects,\nso that the same BAM files can be used in multiple datasets without having to\nbe remade.\u003c/p\u003e\n\u003cp\u003eSo, say you have \u003ccode\u003edataset1_samples.tsv\u003c/code\u003e and \u003ccode\u003edataset2_samples.tsv\u003c/code\u003e, with\ncorresponding \u003ccode\u003edataset1_config.tsv\u003c/code\u003e and \u003ccode\u003edataset2_config.yaml\u003c/code\u003e. The sample\nfiles contain different samples, though some are shared between the datasets.\nThe workflow for dataset1 can be run, and then dataset2 can be run. When\ndataset2 runs, it map new samples, but won\u0027t re-map samples processed in\ndataset1. Each will perform downstream analyses independently with their sample\nset and configuration files, storing these results in dataset specific folders.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch4 class=\"heading-element\"\u003eReference Configuration\u003c/h4\u003e\u003ca id=\"user-content-reference-configuration\" class=\"anchor\" aria-label=\"Permalink: Reference Configuration\" href=\"#reference-configuration\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eRequired configuration of the reference.\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cp\u003e\u003ccode\u003echunk_size:\u003c/code\u003e A size in bp (integer). Your reference will be analyzed in\n\u0027chunks\u0027 of contigs of this size to parallelize processing. This size should\nbe larger than the largest contig in your genome. A larger number means fewer\njobs that run longer. A smaller number means more jobs that run shorter. The\nbest fit will depend on the reference and the compute resources you have\navailable. Leaving this blank will not divide the reference up into chunks\n(but this isn\u0027t optimized yet, so it will do a couple unnecessary steps).\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003ccode\u003ereference:\u003c/code\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003ename:\u003c/code\u003e A name for your reference genome, will go in the file names.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003efasta:\u003c/code\u003e A path to the reference fasta file (currently only supports\nuncompressed fasta files)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003emito:\u003c/code\u003e Mitochondrial contig name(s), will be removed from analysis. Should\nbe listed within brackets []\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003esex-linked:\u003c/code\u003e Sex-linked contig name(s), will be removed from analysis.\nShould be listed within brackets []\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eexclude:\u003c/code\u003e Additional contig name(s) to exclude from analysis. Should be\nlisted within brackets []\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003emin_size:\u003c/code\u003e A size in bp (integer). All contigs below this size will be\nexcluded from analysis.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eReference genomes should be uncompressed, and contig names should be clear and\nconcise. Currently, there are some issues parsing contig names with\nunderscores, so please change these in your reference before running the\npipeline. Alphanumeric characters, as well as \u003ccode\u003e.\u003c/code\u003e in contig names have been\ntested to work so far, other symbols have not been tested.\u003c/p\u003e\n\u003cp\u003ePotentially the ability to use bgzipped genomes will be added, I just need to\ncheck that it works with all underlying tools. Currently, it will for sure not\nwork, as calculating chunks is hard-coded to work on an uncompressed genome.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch4 class=\"heading-element\"\u003eSample Set Configuration\u003c/h4\u003e\u003ca id=\"user-content-sample-set-configuration\" class=\"anchor\" aria-label=\"Permalink: Sample Set Configuration\" href=\"#sample-set-configuration\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eThis will exclude individuals from analysis that are listed in the sample list.\nThis may be useful if you run the workflow and find a poor quality sample, and\nwant to re-run without it. Or if you have relatives in the dataset and you want\nto exclude them where necessary.\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003eexclude_ind:\u003c/code\u003e Sample name(s) that will be excluded from the workflow. Should\nbe a list in [].\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eexcl_pca-admix:\u003c/code\u003e Sample name(s) that will be excluded \u003cem\u003eonly\u003c/em\u003e from PCA and\nAdmixture analyses. Useful for close relatives that violate the assumptions\nof these analyses, but that you want in others. Should be a list in [].\u003c/li\u003e\n\u003c/ul\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch4 class=\"heading-element\"\u003eAnalysis Selection\u003c/h4\u003e\u003ca id=\"user-content-analysis-selection\" class=\"anchor\" aria-label=\"Permalink: Analysis Selection\" href=\"#analysis-selection\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eHere, you will define which analyses you will perform. It is useful to start\nwith only a few, and add more in subsequent workflow runs, just to ensure you\ncatch errors before you use compute time running all analyses. Most are set\nwith (\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e) or a value, described below. Modifications to the\nsettings for each analysis are set in the next section.\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cp\u003e\u003ccode\u003epopulations:\u003c/code\u003e A list of populations found in your sample list to limit\npopulation analyses to. Might be useful if you want to perform individual\nanalyses on some samples but not include them in any population level\nanalyses\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003ccode\u003eanalyses:\u003c/code\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003egenmap:\u003c/code\u003e Filter out sites with low mappability estimated by Genmap\n(\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003erepeatmasker:\u003c/code\u003e (NOTE: Only one of the three options should be filled/true)\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003elocal_lib:\u003c/code\u003e Filter repeats by masking with an already made library you\nhave locally. Should be file path.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003edfam_lib:\u003c/code\u003e Filter repeats using a library available from dfam. Should be\na taxonomic group name.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003ebuild_lib:\u003c/code\u003e Use RepeatModeler to build a library of repeats from the\nreference itself, then filter them from analysis (\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e).\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eextreme_depth:\u003c/code\u003e Filter out sites with extremely high or low global\nsequencing depth (\u003ccode\u003e[lower, upper]\u003c/code\u003e). The value of \u003ccode\u003elower\u003c/code\u003e (float) will be\nmultiplied by the median global depth to create a lower global depth\nthreshold, and \u003ccode\u003eupper\u003c/code\u003e will do the same to creat an upper threshold. This\nis done for all samples, as well as separately for depth groupings defined\nin samples file.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003edataset_missing_data:\u003c/code\u003e A floating point value between 0 and 1. Sites with\ndata for fewer than this proportion of individuals across the whole dataset\nwill be filtered out.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003epopulation_missing_data:\u003c/code\u003e A floating point value between 0 and 1. Sites\nwith data for fewer than this proportion of individuals in any population\nwill be filtered out for all populations.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003equalimap:\u003c/code\u003e Perform Qualimap bamqc on bam files for general quality stats\n(\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003edamageprofiler:\u003c/code\u003e Estimate post-mortem DNA damage on historical samples\nwith Damageprofiler (\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e) NOTE: This just adds the addition of\nDamageprofiler to the already default output of MapDamage. DNA damage will\nalways be estimated and rescaled by MapDamage for samples marked as\n\u0027historical\u0027\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eestimate_ld:\u003c/code\u003e Estimate pairwise linkage disquilibrium between sites with\nngsLD for each popualation and the whole dataset. Note, only set this if\nyou want to generate the LD estimates for use in downstream analyses\noutside this workflow. Other analyses within this workflow that require LD\nestimates (LD decay/pruning) will function properly regardless of the\nsetting here. (\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eld_decay:\u003c/code\u003e Use ngsLD to plot LD decay curves for each population and for\nthe dataset as a whole (\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003epca_pcangsd:\u003c/code\u003e Perform Principal Component Analysis with PCAngsd\n(\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eadmix_ngsadmix:\u003c/code\u003e Perform admixture analysis with NGSadmix (\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003erelatedness:\u003c/code\u003e Can be performed multiple ways, set any combination of the\nthree options below. Note, that I\u0027ve mostly incorporated these with the\nR0/R1/KING kinship methods in Waples et al. 2019, \u003cem\u003eMol. Ecol.\u003c/em\u003e in mind.\nThese methods differ slightly from how they implement this method, and will\ngive slightly more/less accurate estimates of kinship depending on your\nreference\u0027s relationship to your samples. \u003ccode\u003eibsrelate_ibs\u003c/code\u003e uses the\nprobabilities of all possible genotypes, so should be the most accurate\nregardless, but can use a lot of memory and take a long time with many\nsamples. \u003ccode\u003eibsrelate_sfs\u003c/code\u003e is a bit more efficient, as it does things in a\npairwise fashion in parallel, but may be biased if the segregating alleles\nin your populations are not represented in the reference. \u003ccode\u003engsrelate\u003c/code\u003e uses\nseveral methods, one of which is similar to \u003ccode\u003eibsrelate_sfs\u003c/code\u003e, but may be\nless accurate due to incorporating in less data. In my experience,\nNGSrelate is suitable to identify up to third degree relatives in the\ndataset, but only if the exact relationship can be somewhat uncertain (i.e.\nyou don\u0027t need to know the difference between, say, parent/offspring and\nfull sibling pairs, or between second degree and third degree relatives).\nIBSrelate_sfs can get you greater accuracy, but may erroneously inflate\nkinship if your datset has many alleles not represented in your reference.\nIf you notice, for instance, a large number of third degree relatives\n(KING ~0.03 - 0.07) in your dataset that is not expected, it may be worth\ntrying the IBS based method (\u003ccode\u003eibsrelate_ibs\u003c/code\u003e).\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003engsrelate:\u003c/code\u003e Co-estimate inbreeding and pairwise relatedness with\nNGSrelate (\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eibsrelate_ibs:\u003c/code\u003e Estimate pairwise relatedness with the IBS based method\nfrom Waples et al. 2019, \u003cem\u003eMol. Ecol.\u003c/em\u003e. This can use a lot of memory, as\nit has genotype likelihoods for all sites from all samples loaded into\nmemory, so it is done per \u0027chunk\u0027, which still takes a lot of time and\nmemory. (\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eibsrelate_sfs:\u003c/code\u003e Estimate pairwise relatedness with the SFS based method\nfrom Waples et al. 2019, \u003cem\u003eMol. Ecol.\u003c/em\u003e. Enabling this can greatly increase\nthe time needed to build the workflow DAG if you have many samples. As a\nform of this method is implemented in NGSrelate, it may be more\nefficient to only enable that. (\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e)\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003ethetas_angsd:\u003c/code\u003e Estimate pi, theta, and Tajima\u0027s D for each population in\nwindows across the genome using ANGSD (\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eheterozygosity_angsd:\u003c/code\u003e Estimate individual genome-wide heterozygosity\nusing ANGSD (\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003efst_angsd:\u003c/code\u003e Estimate pairwise $F_{ST}$ using ANGSD. Set one or both of the\nbelow options. Estimates both globally and in windows across the genome.\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003epopulations:\u003c/code\u003e Pairwise $F_{ST}$ is calculated between all possible\npopulation pairs (\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eindividuals:\u003c/code\u003e Pairwise $F_{ST}$ is calculated between all possible\npopulation pairs. NOTE: This can be really intensive on the DAG building\nprocess, so I don\u0027t recommend enabling unless you\u0027re certain you want\nthis (\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e)\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003einbreeding_ngsf-hmm:\u003c/code\u003e Estimates inbreeding coefficients and runs of\nhomozygosity using ngsF-HMM. Output is converted into an inbreeding measure\n$F_ROH$, which describes the proportion of the genome in runs of\nhomozygosity over a certain length. (\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eibs_matrix:\u003c/code\u003e Estimate pairwise identity by state distance between all\nsamples using ANGSD. (\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e)\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch4 class=\"heading-element\"\u003eFilter Sets\u003c/h4\u003e\u003ca id=\"user-content-filter-sets\" class=\"anchor\" aria-label=\"Permalink: Filter Sets\" href=\"#filter-sets\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eBy default, this workflow will perform all analyses requested in the above\nsection on all sites that pass the filters set in the above section. These\noutputs will contain \u003ccode\u003eallsites-filts\u003c/code\u003e in the filename and in the report.\nHowever, many times, it is useful to perform an analysis on different subsets\nof sites, for instance, to compare results for genic vs. intergenic regions,\nneutral sites, exons vs. introns, etc. Here, users can set an arbitrary number\nof additional filters using BED files. For each BED file supplied, the contents\nwill be intersected with the sites passing the filters set in the above\nsection, and all analyses will be performed additionally using those sites.\u003c/p\u003e\n\u003cp\u003eFor instance, given a BED file containing putatively neutral sites, one could\nset the following:\u003c/p\u003e\n\u003cdiv class=\"highlight highlight-source-yaml\"\u003e\u003cpre\u003e\u003cspan class=\"pl-ent\"\u003efilter_beds\u003c/span\u003e:\n \u003cspan class=\"pl-ent\"\u003eneutral-sites\u003c/span\u003e: \u003cspan class=\"pl-s\"\u003e\u003cspan class=\"pl-pds\"\u003e\"\u003c/span\u003eresources/neutral_sites.bed\u003cspan class=\"pl-pds\"\u003e\"\u003c/span\u003e\u003c/span\u003e\u003c/pre\u003e\u003c/div\u003e\n\u003cp\u003eIn this case, for each requested analysis, in addition to the \u003ccode\u003eallsites-filts\u003c/code\u003e\noutput, a \u003ccode\u003eneutral-filts\u003c/code\u003e (named after the key assigned to the BED file in\n\u003ccode\u003econfig.yaml\u003c/code\u003e) output will also be generated, containing the results for sites\nwithin the specified BED file that passed any set filters.\u003c/p\u003e\n\u003cp\u003eMore than one BED file can be set, up to an arbitrary number:\u003c/p\u003e\n\u003cdiv class=\"highlight highlight-source-yaml\"\u003e\u003cpre\u003e\u003cspan class=\"pl-ent\"\u003efilter_beds\u003c/span\u003e:\n \u003cspan class=\"pl-ent\"\u003eneutral\u003c/span\u003e: \u003cspan class=\"pl-s\"\u003e\u003cspan class=\"pl-pds\"\u003e\"\u003c/span\u003eresources/neutral_sites.bed\u003cspan class=\"pl-pds\"\u003e\"\u003c/span\u003e\u003c/span\u003e\n \u003cspan class=\"pl-ent\"\u003eintergenic\u003c/span\u003e: \u003cspan class=\"pl-s\"\u003e\u003cspan class=\"pl-pds\"\u003e\"\u003c/span\u003eresources/intergenic_sites.bed\u003cspan class=\"pl-pds\"\u003e\"\u003c/span\u003e\u003c/span\u003e\n \u003cspan class=\"pl-ent\"\u003eintrons\u003c/span\u003e: \u003cspan class=\"pl-s\"\u003e\u003cspan class=\"pl-pds\"\u003e\"\u003c/span\u003eresources/introns.bed\u003cspan class=\"pl-pds\"\u003e\"\u003c/span\u003e\u003c/span\u003e\u003c/pre\u003e\u003c/div\u003e\n\u003cp\u003eIt may also sometimes be desireable to skip analyses on \u003ccode\u003eallsites-filts\u003c/code\u003e, say\nif you are trying to only generate diversity estimates or generate SFS for a\nset of neutral sites you supply. To skip running any analyses for\n\u003ccode\u003eallsites-filts\u003c/code\u003e and only perform them for the BED files you supply, you can\nset \u003ccode\u003eonly_filter_beds: true\u003c/code\u003e in the config file. This may also be useful in the\nevent you have a set of already filtered sites, and want to run the workflow on\nthose, ignoring any of the built in filter options by setting them to \u003ccode\u003efalse\u003c/code\u003e.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch4 class=\"heading-element\"\u003eSoftware Configuration\u003c/h4\u003e\u003ca id=\"user-content-software-configuration\" class=\"anchor\" aria-label=\"Permalink: Software Configuration\" href=\"#software-configuration\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eThese are software specific settings that can be user configured in the\nworkflow. If you are missing a configurable setting you need, open up an issue\nor a pull request and I\u0027ll gladly put it in.\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cp\u003e\u003ccode\u003emapQ:\u003c/code\u003e Phred-scaled mapping quality filter. Reads below this threshold will\nbe filtered out. (integer)\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003ccode\u003ebaseQ:\u003c/code\u003e Phred-scaled base quality filter. Reads below this threshold will be\nfiltered out. (integer)\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003ccode\u003eparams:\u003c/code\u003e\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003egenmap:\u003c/code\u003e Parameters for mappability analysis, see \u003ca href=\"https://github.com/cpockrandt/genmap/\"\u003eGenMap\u0027s documentation\u003c/a\u003e\nfor more details.\n\u003cul\u003e\n\u003cli\u003e\u003ccode\u003eK:\u003c/code\u003e\u003c/li\u003e\n\u003cli\u003e\u003ccode\u003eE:\u003c/code\u003e\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003emap_thresh:\u003c/code\u003e A threshold mappability score. Sites with a mappability\nscore below this threshold are filtered out if GenMap is enabled.\n(integer/float, 0-1)\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003efastp:\u003c/code\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003eextra:\u003c/code\u003e Additional options to pass to fastp trimming. (string)\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003epicard:\u003c/code\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003eMarkDuplicates:\u003c/code\u003e Additional options to pass to Picard MarkDuplicates.\n\u003ccode\u003e--REMOVE_DUPLICATES true\u003c/code\u003e is recommended. (string)\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eangsd:\u003c/code\u003e General options in ANGSD, relevant doc pages are linked\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003egl_model:\u003c/code\u003e Genotype likelihood model to use in calculation\n(\u003ccode\u003e-GL\u003c/code\u003e option in ANGSD, \u003ca href=\"http://www.popgen.dk/angsd/index.php/Genotype_Likelihoods\" rel=\"nofollow\"\u003edocs\u003c/a\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003emaxdepth:\u003c/code\u003e When calculating individual depth, sites with depth higher\nthan this will be binned to this value. Should be fine for most to leave\nat \u003ccode\u003e1000\u003c/code\u003e. (integer, \u003ca href=\"http://www.popgen.dk/angsd/index.php/Depth\" rel=\"nofollow\"\u003edocs\u003c/a\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eextra:\u003c/code\u003e Additional options to pass to ANGSD during genotype likelihood\ncalculation. This is primarily useful for adding BAM input filters. Note\nthat \u003ccode\u003e--remove_bads\u003c/code\u003e and \u003ccode\u003e-only_proper_pairs\u003c/code\u003e are enabled by default, so\nthey only need to be included if you want to turn them off. I\u0027ve also\nfound that for some datasets, \u003ccode\u003e-C 50\u003c/code\u003e and \u003ccode\u003e-baq 1\u003c/code\u003e can create a strong\nrelationship between sample depth and detected diversity, effectively\nremoving the benefits of ANGSD for low/variable depth data. I recommend\nthat these aren\u0027t included unless you know you need them, and even then,\nI\u0027d recommend plotting \u003ccode\u003eheterozygosity ~ sample depth\u003c/code\u003e to ensure there is\nnot any relationship. Since the workflow uses bwa to map, \u003ccode\u003e-uniqueOnly 1\u003c/code\u003e\ndoesn\u0027t do anything if your minimum mapping quality is \u0026gt; 0. Don\u0027t put\nmapping and base quality thresholds here either, it will use the ones\ndefined above automatically. Although historical samples will have DNA\ndamaged assessed and to some extent, corrected, it may be useful to put\n\u003ccode\u003e-noTrans 1\u003c/code\u003e or \u003ccode\u003e-trim INT\u003c/code\u003e here if you\u0027re interested in stricter filters\nfor degraded DNA. (string, \u003ca href=\"http://www.popgen.dk/angsd/index.php/Input#BAM.2FCRAM\" rel=\"nofollow\"\u003edocs\u003c/a\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003esnp_pval:\u003c/code\u003e The p-value to use for calling SNPs (float, \u003ca href=\"http://www.popgen.dk/angsd/index.php/SNP_calling\" rel=\"nofollow\"\u003edocs\u003c/a\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003emin_maf:\u003c/code\u003e The minimum minor allele frequency required to call a SNP.\n(float, \u003ca href=\"http://www.popgen.dk/angsd/index.php/Allele_Frequencies\" rel=\"nofollow\"\u003edocs\u003c/a\u003e)\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003engsld:\u003c/code\u003e Settings for ngsLD (\u003ca href=\"https://github.com/fgvieira/ngsLD\"\u003edocs\u003c/a\u003e)\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003emax_kb_dist_est-ld:\u003c/code\u003e For the LD estimates generated when setting\n\u003ccode\u003eestimate_ld: true\u003c/code\u003e above, set the maximum distance between sites in kb\nthat LD will be estimated for (\u003ccode\u003e--max_kb_dist\u003c/code\u003e in ngsLD, integer)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003emax_kb_dist_decay:\u003c/code\u003e The same as \u003ccode\u003emax_kb_dist_est-ld:\u003c/code\u003e, but used when\nestimating LD decay when setting \u003ccode\u003eld_decay: true\u003c/code\u003e above (integer)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003emax_kb_dist_pruning:\u003c/code\u003e The same as \u003ccode\u003emax_kb_dist_est-ld:\u003c/code\u003e, but used when\nlinkage pruning SNPs as inputs for PCA, Admix, and Inbreeding analyses.\nAny positions above this distance will be assumed to be in linkage\nequilibrium during the pruning process (integer)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003ernd_sample_est-ld:\u003c/code\u003e For the LD estimates generated when setting\n\u003ccode\u003eestimate_ld: true\u003c/code\u003e above, randomly sample this proportion of pairwise\nlinkage estimates rather than estimating all (\u003ccode\u003e--rnd_sample\u003c/code\u003e in ngsLD,\nfloat)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003ernd_sample_decay:\u003c/code\u003e The same as \u003ccode\u003ernd_sample_est-ld:\u003c/code\u003e, but used when\nestimating LD decay when setting \u003ccode\u003eld_decay: true\u003c/code\u003e above (float)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003efit_LDdecay_extra:\u003c/code\u003e Additional plotting arguments to pass to\n\u003ccode\u003efit_LDdecay.R\u003c/code\u003e when estimating LD decay (string)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003efit_LDdecay_n_correction:\u003c/code\u003e When estimating LD decay, should the sample\nsize corrected r^2 model be used? (\u003ccode\u003etrue\u003c/code\u003e/\u003ccode\u003efalse\u003c/code\u003e, \u003ccode\u003etrue\u003c/code\u003e is the\nequivalent of passing a sample size to \u003ccode\u003efit_LDdecay.R\u003c/code\u003e in ngsLD using\n\u003ccode\u003e--n_ind\u003c/code\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003epruning_min-weight:\u003c/code\u003e The minimum r^2 to assume two positions are in\nlinkage disequilibrium when pruning (float)\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003erealSFS:\u003c/code\u003e Settings for realSFS\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003efold:\u003c/code\u003e Whether or not to fold the produced SFS (0 or 1, \u003ca href=\"http://www.popgen.dk/angsd/index.php/SFS_Estimation\" rel=\"nofollow\"\u003edocs\u003c/a\u003e)\n\u003cstrong\u003eNOTE:\u003c/strong\u003e I have not implemented the use of an ancestral reference into\nthis workflow, so this should always be set to 1 until I implement this.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003esfsboot:\u003c/code\u003e Doesn\u0027t work now, but when it does it will produce this many\nbootstrapped SFS per population and population pair (integer)\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003efst:\u003c/code\u003e Settings for $F_{ST}$ calculation in ANGSD\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003ewhichFst:\u003c/code\u003e Determines which $F_{ST}$ estimator is used by ANGSD. With 0\nbeing the default Reynolds 1983 and 1 being the Bhatia 2013 estimator.\nThe latter is preferable for small or uneven sample sizes\n(0 or 1, \u003ca href=\"http://www.popgen.dk/angsd/index.php/Fst\" rel=\"nofollow\"\u003edocs\u003c/a\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003ewin_size:\u003c/code\u003e Window size in bp for sliding window analysis (integer)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003ewin_step:\u003c/code\u003e Window step size in bp for sliding window analysis (integer)\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003ethetas:\u003c/code\u003e Settings for pi, theta, and Tajima\u0027s D estimation\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003ewin_size:\u003c/code\u003e Window size in bp for sliding window analysis (integer)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003ewin_step:\u003c/code\u003e Window step size in bp for sliding window analysis (integer)\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003engsadmix:\u003c/code\u003e Settings for admixture analysis with NGSadmix. This analysis is\nperformed for a set of K groupings, and each K has several replicates\nperformed. Replicates will continue until a set of N highest likelihood\nreplicates converge, or the number of replicates reaches an upper threshold\nset here. Defaults for \u003ccode\u003ereps\u003c/code\u003e, \u003ccode\u003eminreps\u003c/code\u003e, \u003ccode\u003ethresh\u003c/code\u003e, and \u003ccode\u003econv\u003c/code\u003e can be left\nas default for most.\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003ekvalues:\u003c/code\u003e A list of values of K to fit the data to (list of integers)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003ereps:\u003c/code\u003e The maximum number of replicates to perform per K (integer)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eminreps:\u003c/code\u003e The minimum number of replicates to perform, even if\nreplicates have converged (integer)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003ethresh:\u003c/code\u003e The convergence threshold - the top replicates must all be\nwithin this value of log-likelihood units to consider the run converged\n(integer)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003econv:\u003c/code\u003e The number of top replicates to include in convergence\nassessment. (integer)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eextra:\u003c/code\u003e Additional arguments to pass to NGSadmix (for instance,\nincreasing \u003ccode\u003e-maxiter\u003c/code\u003e). (string, \u003ca href=\"http://www.popgen.dk/software/index.php/NgsAdmix\" rel=\"nofollow\"\u003edocs\u003c/a\u003e)\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eibs:\u003c/code\u003e Settings for identity by state calculation with ANGSD\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003e-doIBS:\u003c/code\u003e Whether to use a random (1) or consensus (2) base in IBS\ndistance calculation (\u003ca href=\"http://www.popgen.dk/angsd/index.php/PCA_MDS\" rel=\"nofollow\"\u003edocs\u003c/a\u003e)\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n", "data_format": 2, @@ -50927,6 +50978,62 @@ var data = ], "updated_at": 1602644548.0 }, + { + "config_readme": null, + "data_format": 2, + "description": "Multi-omics pipeline for the GLAMR database", + "formatting": "[DEBUG] \n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"input\" at line 458 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"input\" at line 797 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"shell\" at line 803 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"input\" at line 960 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"output\" at line 964 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"output\" at line 1017 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"output\" at line 1036 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"input\" at line 1056 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"output\" at line 1060 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"output\" at line 1112 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"output\" at line 1644 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"output\" at line 1710 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"input\" at line 2008 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"input\" at line 2055 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"input\" at line 2176 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"output\" at line 2215 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"output\" at line 2659 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"input\" at line 2881 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"input\" at line 2893 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"params\" at line 2917 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"params\" at line 3017 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmp9ubijvwd/Snakefile\": Keyword \"shell\" at line 3057 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[ERROR] In file \"/tmp/tmp9ubijvwd/Snakefile\": NoParametersError: L3116: In params definition.\n[INFO] In file \"/tmp/tmp9ubijvwd/Snakefile\": 1 file(s) raised parsing errors \ud83e\udd15\n\nsnakefmt version: 0.10.2", + "full_name": "Geo-omics/GLAMR_omics_pipelines", + "latest_release": null, + "linting": "/tmp/tmp9ubijvwd/Snakefile:6437: SyntaxWarning: invalid escape sequence \u0027\\;\u0027\n/tmp/tmp9ubijvwd/Snakefile:6458: SyntaxWarning: invalid escape sequence \u0027\\;\u0027\nFileNotFoundError in file /tmp/tmp9ubijvwd/Snakefile, line 35:\n[Errno 2] No such file or directory: \u0027data/sample_metadata/sample_lists/metaG_samples\u0027\n File \"/tmp/tmp9ubijvwd/Snakefile\", line 35, in \u003cmodule\u003e\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 6, + "subscribers_count": 4, + "topics": [ + "noaa-omics-software" + ], + "updated_at": 1726031221.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "The Loaders for GraphKB. Imports content from external sources via the GraphKB REST API", + "formatting": "[DEBUG] \n\u003cunknown\u003e:7: SyntaxWarning: invalid escape sequence \u0027\\/\u0027\n\u003cunknown\u003e:10: SyntaxWarning: invalid escape sequence \u0027\\ \u0027\n[ERROR] In file \"/tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile\": NoParametersError: L399: In input definition.\n[INFO] In file \"/tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile\": 1 file(s) raised parsing errors \ud83e\udd15\n\nsnakefmt version: 0.10.2", + "full_name": "bcgsc/pori_graphkb_loader", + "latest_release": "v8.0.2", + "linting": "/tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile:197: SyntaxWarning: invalid escape sequence \u0027\\ \u0027\n cd {DATA_DIR}/cosmic\n/tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile:206: SyntaxWarning: invalid escape sequence \u0027\\ \u0027\n rule load_local:\nLints for snakefile /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile:\n * Absolute path \"/\u003e\u003cAttribute name = \" in line 69:\n Do not define absolute paths inside of the workflow, since this renders\n your workflow irreproducible on other machines. Use path relative to the\n working directory instead, or make the path configurable via a config\n file.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/configuration.html#configuration\n * Absolute path \"/releases/[^\" in line 118:\n Do not define absolute paths inside of the workflow, since this renders\n your workflow irreproducible on other machines. Use path relative to the\n working directory instead, or make the path configurable via a config\n file.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/configuration.html#configuration\n * Mixed rules and functions in same snakefile.:\n Small one-liner functions used only once should be defined as lambda\n expressions. Other functions should be collected in a common module, e.g.\n \u0027rules/common.smk\u0027. This makes the workflow steps more readable.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/modularization.html#includes\n * Path composition with \u0027+\u0027 in line 211:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 220:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 229:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 239:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 248:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 257:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 266:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 275:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 284:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 292:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 333:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 344:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 353:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 363:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 373:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 384:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 394:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 402:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 412:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 424:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 434:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 443:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 459:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n\nLints for rule download_ncit (line 48, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_ncit_fda (line 58, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_ensembl (line 65, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_fda_srs (line 74, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_refseq (line 86, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_uberon (line 94, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_do (line 102, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_drugbank (line 113, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable filename from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable filename from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule download_PMC4468049 (line 127, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_PMC4232638 (line 135, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_cgi (line 143, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_local_data (line 152, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_cancerhotspots (line 160, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_cosmic_resistance (line 171, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_cosmic_diseases (line 183, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_cosmic_fusions (line 194, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule all_drugs (line 304, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule all_diseases (line 311, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule download_sources (line 447, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule all_ontologies (line 462, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 6, + "subscribers_count": 6, + "topics": [ + "pori", + "etl" + ], + "updated_at": 1726008134.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "An experiment to determine the relationship between spatial frequency and eccentricity in the human early visual cortex.", + "formatting": "[DEBUG] \n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\.\u0027\n[DEBUG] In file \"/tmp/tmpsv5kdkrz/billbrod-spatial-frequency-preferences-e4fa925/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "billbrod/spatial-frequency-preferences", + "latest_release": "v1.1.1", + "linting": "/tmp/tmpsv5kdkrz/billbrod-spatial-frequency-preferences-e4fa925/Snakefile:116: SyntaxWarning: invalid escape sequence \u0027\\.\u0027\n task=\"task-[a-z0-9]+\",\n/tmp/tmpsv5kdkrz/billbrod-spatial-frequency-preferences-e4fa925/Snakefile:831: SyntaxWarning: invalid escape sequence \u0027\\*\u0027\n atlases = {}\n/tmp/tmpsv5kdkrz/billbrod-spatial-frequency-preferences-e4fa925/Snakefile:832: SyntaxWarning: invalid escape sequence \u0027\\*\u0027\n for hemi in [\u0027lh\u0027, \u0027rh\u0027]:\nModuleNotFoundError in file /tmp/tmpsv5kdkrz/billbrod-spatial-frequency-preferences-e4fa925/Snakefile, line 6:\nNo module named \u0027neuropythy\u0027\n File \"/tmp/tmpsv5kdkrz/billbrod-spatial-frequency-preferences-e4fa925/Snakefile\", line 6, in \u003cmodule\u003e\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 6, + "subscribers_count": 2, + "topics": [], + "updated_at": 1725978607.0 + }, { "config_readme": "\u003cdiv class=\"markdown-heading\"\u003e\u003ch1 class=\"heading-element\"\u003eDependencies installation\u003c/h1\u003e\u003ca id=\"user-content-dependencies-installation\" class=\"anchor\" aria-label=\"Permalink: Dependencies installation\" href=\"#dependencies-installation\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eCreate and activate the virtual environment with the required dependencies\nwith Conda:\u003c/p\u003e\n\u003cdiv class=\"highlight highlight-source-shell\"\u003e\u003cpre\u003econda env create -f environment.yml\nconda activate mirflowz\u003c/pre\u003e\u003c/div\u003e\n\u003cp\u003eIf you plan to run \u003cem\u003eMIRFLOWZ\u003c/em\u003e via Conda, we recommend using the following\ncommand for a faster environment creation, specially if you will run it on an\nHPC cluster.\u003c/p\u003e\n\u003cdiv class=\"highlight highlight-source-shell\"\u003e\u003cpre\u003econda config --set channel_priority strict\u003c/pre\u003e\u003c/div\u003e\n\u003cp\u003eFor a faster creation of the environment (and Conda environments in general),\nyou can also install \u003ca href=\"https://github.com/mamba-org/mamba\"\u003eMamba\u003c/a\u003e on top of Conda. In that case, replace\n\u003ccode\u003econda\u003c/code\u003e with \u003ccode\u003emamba\u003c/code\u003e in the commands above (particularly in\n\u003ccode\u003econda env create\u003c/code\u003e).\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch2 class=\"heading-element\"\u003eRunning \u003cem\u003eMIRFLOWZ\u003c/em\u003e with Singularity\u003c/h2\u003e\u003ca id=\"user-content-running-mirflowz-with-singularity\" class=\"anchor\" aria-label=\"Permalink: Running MIRFLOWZ with Singularity\" href=\"#running-mirflowz-with-singularity\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eIf you want to run \u003cem\u003eMIRFLOWZ\u003c/em\u003e via Singularity and do not already\nhave it installed globally on your system, you must further update the Conda\nenvironment with:\u003c/p\u003e\n\u003cdiv class=\"highlight highlight-source-shell\"\u003e\u003cpre\u003econda env update -f environment.root.yml\u003c/pre\u003e\u003c/div\u003e\n\u003cblockquote\u003e\n\u003cp\u003eMind that you must have the environment activated and root permissions on\nyour system to install Singularity. If you want to run \u003cem\u003eMIRFLOWZ\u003c/em\u003e on an HPC\ncluster (recommended in almost all cases), ask your system administrator\nabout Singularity.\u003c/p\u003e\n\u003c/blockquote\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch1 class=\"heading-element\"\u003eRun the workflow on your own samples\u003c/h1\u003e\u003ca id=\"user-content-run-the-workflow-on-your-own-samples\" class=\"anchor\" aria-label=\"Permalink: Run the workflow on your own samples\" href=\"#run-the-workflow-on-your-own-samples\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eIn order to run \u003cem\u003eMIRFLOWZ\u003c/em\u003e on your own samples, we recommend having all the\ninput files inside a dedicated directory. This way, it is easier to keep the\ndata together and reproduce an analysis. Assuming that your current directory\nis the repository\u0027s root directory, create a directory to store all your data\nand traverse to it with:\u003c/p\u003e\n\u003cdiv class=\"highlight highlight-source-shell\"\u003e\u003cpre\u003emkdir path/to/your_run\n\u003cspan class=\"pl-c1\"\u003ecd\u003c/span\u003e path/to/your_run\u003c/pre\u003e\u003c/div\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch2 class=\"heading-element\"\u003e1. Prepare the sample table\u003c/h2\u003e\u003ca id=\"user-content-1-prepare-the-sample-table\" class=\"anchor\" aria-label=\"Permalink: 1. Prepare the sample table\" href=\"#1-prepare-the-sample-table\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eCreate an empty sample table. Refer to the\n\u003ca href=\"../test/test_files/samples_table.tsv\"\u003esample.tsv\u003c/a\u003e test file to see what the\ntable must look like or use it as a template.\u003c/p\u003e\n\u003cdiv class=\"highlight highlight-source-shell\"\u003e\u003cpre\u003etouch samples.tsv\u003c/pre\u003e\u003c/div\u003e\n\u003cblockquote\u003e\n\u003cp\u003eFill the sample table according to the following requirements:\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003esample\u003c/code\u003e. Arbitrary name for the miRNA sequencing library.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003esample_file\u003c/code\u003e. Path to the miRNA sequencing library file. The path must be\nrelative to the directory where the workflow will be run.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eadapter\u003c/code\u003e. Sequence of the 3\u0027-end adapter used during library preparation.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003eformat\u003c/code\u003e. One of \u003ccode\u003efa\u003c/code\u003e/\u003ccode\u003efasta\u003c/code\u003e or \u003ccode\u003efq\u003c/code\u003e/\u003ccode\u003efastq\u003c/code\u003e, if the library file is in\nFASTA or FASTQ format, respectively.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/blockquote\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch2 class=\"heading-element\"\u003e2. Prepare the genome resources\u003c/h2\u003e\u003ca id=\"user-content-2-prepare-the-genome-resources\" class=\"anchor\" aria-label=\"Permalink: 2. Prepare the genome resources\" href=\"#2-prepare-the-genome-resources\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eThere are 4 files you must provide:\u003c/p\u003e\n\u003col\u003e\n\u003cli\u003e\n\u003cp\u003eA \u003cstrong\u003e\u003ccode\u003egzip\u003c/code\u003eped FASTA\u003c/strong\u003e file containing \u003cstrong\u003ereference sequences\u003c/strong\u003e, typically the\ngenome of the source/organism from which the library was extracted.\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003eA \u003cstrong\u003e\u003ccode\u003egzip\u003c/code\u003eped GTF\u003c/strong\u003e file with matching \u003cstrong\u003egene annotations\u003c/strong\u003e for the\nreference sequences above.\u003c/p\u003e\n\u003c/li\u003e\n\u003c/ol\u003e\n\u003cblockquote\u003e\n\u003cp\u003e\u003cem\u003eMIRFLOWZ\u003c/em\u003e expects both the reference sequence and gene annotation files to\nfollow \u003ca href=\"https://ensembl.org/\" rel=\"nofollow\"\u003eEnsembl\u003c/a\u003e style/formatting. If you obtained these files from\na source other than Ensembl, you must ensure that they adhere to the\nexpected format by converting them, if necessary.\u003c/p\u003e\n\u003c/blockquote\u003e\n\u003col start=\"3\"\u003e\n\u003cli\u003eAn \u003cstrong\u003euncompressed GFF3\u003c/strong\u003e file with \u003cstrong\u003emicroRNA annotations\u003c/strong\u003e for the reference\nsequences above.\u003c/li\u003e\n\u003c/ol\u003e\n\u003cblockquote\u003e\n\u003cp\u003e\u003cem\u003eMIRFLOWZ\u003c/em\u003e expects the miRNA annotations to follow \u003ca href=\"https://mirbase.org/\" rel=\"nofollow\"\u003emiRBase\u003c/a\u003e\nstyle/formatting. If you obtained this file from a source other than miRBase,\nyou must ensure that it adheres to the expected format by converting it, if\nnecessary.\u003c/p\u003e\n\u003c/blockquote\u003e\n\u003col start=\"4\"\u003e\n\u003cli\u003e\n\u003cp\u003eAn \u003cstrong\u003euncompressed tab-separated file\u003c/strong\u003e with a \u003cstrong\u003emapping between the\nreference names\u003c/strong\u003e used in the miRNA annotation file (column 1; \"UCSC style\")\nand in the gene annotations and reference sequence files (column 2; \"Ensembl\nstyle\"). Values in column 1 are expected to be unique, no header is\nexpected, and any additional columns will be ignored. \u003ca href=\"https://github.com/dpryan79/ChromosomeMappings\"\u003eThis\nresource\u003c/a\u003e provides such files for various organisms, and in the\nexpected format.\u003c/p\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003cp\u003e\u003cstrong\u003eOPTIONAL\u003c/strong\u003e: A \u003cstrong\u003eBED6\u003c/strong\u003e file with regions for which to produce\n\u003ca href=\"https://git.scicore.unibas.ch/zavolan_group/tools/ascii-alignment-pileup\" rel=\"nofollow\"\u003eASCII-style alignment pileups\u003c/a\u003e. If not provided, no pileups\nwill be generated. See \u003ca href=\"https://gist.github.com/deliaBlue/19ad3740c95937378bd9281bd9d1bc72\"\u003ehere\u003c/a\u003e for the expected format.\u003c/p\u003e\n\u003c/li\u003e\n\u003c/ol\u003e\n\u003cblockquote\u003e\n\u003cp\u003eGeneral note: If you want to process the genome resources before use (e.g.,\nfiltering), you can do that, but make sure the formats of any modified\nresource files meet the formatting expectations outlined above!\u003c/p\u003e\n\u003c/blockquote\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch2 class=\"heading-element\"\u003e3. Prepare the configuration file\u003c/h2\u003e\u003ca id=\"user-content-3-prepare-the-configuration-file\" class=\"anchor\" aria-label=\"Permalink: 3. Prepare the configuration file\" href=\"#3-prepare-the-configuration-file\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eWe recommend creating a copy of the\n\u003ca href=\"config_template.yaml\"\u003econfiguration file template\u003c/a\u003e.\u003c/p\u003e\n\u003cdiv class=\"highlight highlight-source-shell\"\u003e\u003cpre\u003ecp ../config/config_template.yaml config.yaml\n\u003c/pre\u003e\u003c/div\u003e\n\u003cp\u003eOpen the new copy in your editor of choice and adjust the configuration\nparameters to your liking. The template explains what each of the parameters\nmean and how you can meaningfully adjust them.\u003c/p\u003e\n", "data_format": 2, @@ -50987,62 +51094,6 @@ var data = "topics": [], "updated_at": 1725434923.0 }, - { - "config_readme": null, - "data_format": 2, - "description": "The Loaders for GraphKB. Imports content from external sources via the GraphKB REST API", - "formatting": "[DEBUG] \n\u003cunknown\u003e:7: SyntaxWarning: invalid escape sequence \u0027\\/\u0027\n\u003cunknown\u003e:10: SyntaxWarning: invalid escape sequence \u0027\\ \u0027\n[ERROR] In file \"/tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile\": NoParametersError: L399: In input definition.\n[INFO] In file \"/tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile\": 1 file(s) raised parsing errors \ud83e\udd15\n\nsnakefmt version: 0.10.2", - "full_name": "bcgsc/pori_graphkb_loader", - "latest_release": "v8.0.2", - "linting": "/tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile:197: SyntaxWarning: invalid escape sequence \u0027\\ \u0027\n cd {DATA_DIR}/cosmic\n/tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile:206: SyntaxWarning: invalid escape sequence \u0027\\ \u0027\n rule load_local:\nLints for snakefile /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile:\n * Absolute path \"/\u003e\u003cAttribute name = \" in line 69:\n Do not define absolute paths inside of the workflow, since this renders\n your workflow irreproducible on other machines. Use path relative to the\n working directory instead, or make the path configurable via a config\n file.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/configuration.html#configuration\n * Absolute path \"/releases/[^\" in line 118:\n Do not define absolute paths inside of the workflow, since this renders\n your workflow irreproducible on other machines. Use path relative to the\n working directory instead, or make the path configurable via a config\n file.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/configuration.html#configuration\n * Mixed rules and functions in same snakefile.:\n Small one-liner functions used only once should be defined as lambda\n expressions. Other functions should be collected in a common module, e.g.\n \u0027rules/common.smk\u0027. This makes the workflow steps more readable.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/modularization.html#includes\n * Path composition with \u0027+\u0027 in line 211:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 220:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 229:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 239:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 248:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 257:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 266:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 275:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 284:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 292:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 333:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 344:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 353:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 363:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 373:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 384:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 394:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 402:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 412:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 424:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 434:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 443:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n * Path composition with \u0027+\u0027 in line 459:\n This becomes quickly unreadable. Usually, it is better to endure some\n redundancy against having a more readable workflow. Hence, just repeat\n common prefixes. If path composition is unavoidable, use pathlib or\n (python \u003e= 3.6) string formatting with f\"...\".\n\nLints for rule download_ncit (line 48, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_ncit_fda (line 58, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_ensembl (line 65, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_fda_srs (line 74, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_refseq (line 86, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_uberon (line 94, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_do (line 102, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_drugbank (line 113, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n * Shell command directly uses variable filename from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n * Shell command directly uses variable filename from outside of the rule:\n It is recommended to pass all files as input and output, and non-file\n parameters via the params directive. Otherwise, provenance tracking is\n less accurate.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#non-file-parameters-for-rules\n\nLints for rule download_PMC4468049 (line 127, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_PMC4232638 (line 135, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_cgi (line 143, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_local_data (line 152, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_cancerhotspots (line 160, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_cosmic_resistance (line 171, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_cosmic_diseases (line 183, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule download_cosmic_fusions (line 194, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule all_drugs (line 304, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule all_diseases (line 311, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\nLints for rule download_sources (line 447, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\nLints for rule all_ontologies (line 462, /tmp/tmpcjqs1_rr/bcgsc-pori_graphkb_loader-4c25b32/Snakefile):\n * No log directive defined:\n Without a log directive, all output will be printed to the terminal. In\n distributed environments, this means that errors are harder to discover.\n In local environments, output of concurrent jobs will be mixed and become\n unreadable.\n Also see:\n https://snakemake.readthedocs.io/en/stable/snakefiles/rules.html#log-files\n\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 6, - "subscribers_count": 6, - "topics": [ - "pori", - "etl" - ], - "updated_at": 1726008134.0 - }, - { - "config_readme": null, - "data_format": 2, - "description": "An experiment to determine the relationship between spatial frequency and eccentricity in the human early visual cortex.", - "formatting": "[DEBUG] \n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\.\u0027\n[DEBUG] In file \"/tmp/tmpsv5kdkrz/billbrod-spatial-frequency-preferences-e4fa925/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "billbrod/spatial-frequency-preferences", - "latest_release": "v1.1.1", - "linting": "/tmp/tmpsv5kdkrz/billbrod-spatial-frequency-preferences-e4fa925/Snakefile:116: SyntaxWarning: invalid escape sequence \u0027\\.\u0027\n task=\"task-[a-z0-9]+\",\n/tmp/tmpsv5kdkrz/billbrod-spatial-frequency-preferences-e4fa925/Snakefile:831: SyntaxWarning: invalid escape sequence \u0027\\*\u0027\n atlases = {}\n/tmp/tmpsv5kdkrz/billbrod-spatial-frequency-preferences-e4fa925/Snakefile:832: SyntaxWarning: invalid escape sequence \u0027\\*\u0027\n for hemi in [\u0027lh\u0027, \u0027rh\u0027]:\nModuleNotFoundError in file /tmp/tmpsv5kdkrz/billbrod-spatial-frequency-preferences-e4fa925/Snakefile, line 6:\nNo module named \u0027neuropythy\u0027\n File \"/tmp/tmpsv5kdkrz/billbrod-spatial-frequency-preferences-e4fa925/Snakefile\", line 6, in \u003cmodule\u003e\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 6, - "subscribers_count": 2, - "topics": [], - "updated_at": 1725978607.0 - }, - { - "config_readme": null, - "data_format": 2, - "description": "Multi-omics pipeline for the GLAMR database", - "formatting": "[DEBUG] \n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"input\" at line 458 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"input\" at line 797 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"shell\" at line 803 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"input\" at line 960 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"output\" at line 964 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"output\" at line 1017 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"output\" at line 1036 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"input\" at line 1056 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"output\" at line 1060 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"output\" at line 1112 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"output\" at line 1644 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"output\" at line 1710 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"input\" at line 2008 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"input\" at line 2055 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"input\" at line 2176 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"output\" at line 2215 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"output\" at line 2659 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"input\" at line 2881 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"input\" at line 2893 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"params\" at line 2917 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"params\" at line 3017 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpxj33dnph/Snakefile\": Keyword \"shell\" at line 3057 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[ERROR] In file \"/tmp/tmpxj33dnph/Snakefile\": NoParametersError: L3116: In params definition.\n[INFO] In file \"/tmp/tmpxj33dnph/Snakefile\": 1 file(s) raised parsing errors \ud83e\udd15\n\nsnakefmt version: 0.10.2", - "full_name": "Geo-omics/GLAMR_omics_pipelines", - "latest_release": null, - "linting": "/tmp/tmpxj33dnph/Snakefile:6437: SyntaxWarning: invalid escape sequence \u0027\\;\u0027\n/tmp/tmpxj33dnph/Snakefile:6458: SyntaxWarning: invalid escape sequence \u0027\\;\u0027\nTraceback (most recent call last):\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/cli.py\", line 2024, in args_to_api\n any_lint = workflow_api.lint()\n ^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/api.py\", line 337, in _handle_no_dag\n return method(self, *args, **kwargs)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/api.py\", line 354, in lint\n workflow.include(\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/workflow.py\", line 1412, in include\n exec(compile(code, snakefile.get_path_or_uri(), \"exec\"), self.globals)\n\n File \"/tmp/tmpxj33dnph/Snakefile\", line 41, in \u003cmodule\u003e\n glerl_samples = open(\"data/sample_metadata/sample_lists/glerl_samples\").read().splitlines()\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n\nFileNotFoundError: [Errno 2] No such file or directory: \u0027data/sample_metadata/sample_lists/metaG_samples\u0027\n\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 6, - "subscribers_count": 4, - "topics": [ - "noaa-omics-software" - ], - "updated_at": 1725079107.0 - }, { "config_readme": null, "data_format": 2, @@ -53135,6 +53186,28 @@ var data = "topics": [], "updated_at": 1618875925.0 }, + { + "config_readme": null, + "data_format": 2, + "description": "Signaling Pathway Reconstruction Analysis Streamliner (SPRAS)", + "formatting": "[DEBUG] \n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\w\u0027\n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\w\u0027\n[DEBUG] In file \"/tmp/tmpxxvt0ijh/Reed-CompBio-spras-ae3f2cd/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "Reed-CompBio/spras", + "latest_release": "0.2.0", + "linting": "/tmp/tmpxxvt0ijh/Reed-CompBio-spras-ae3f2cd/Snakefile:14: SyntaxWarning: invalid escape sequence \u0027\\w\u0027\n params=\"params-\\w+\",\n/tmp/tmpxxvt0ijh/Reed-CompBio-spras-ae3f2cd/Snakefile:15: SyntaxWarning: invalid escape sequence \u0027\\w\u0027\n dataset=\"\\w+\"\nModuleNotFoundError in file /tmp/tmpxxvt0ijh/Reed-CompBio-spras-ae3f2cd/Snakefile, line 2:\nNo module named \u0027docker\u0027\n File \"/tmp/tmpxxvt0ijh/Reed-CompBio-spras-ae3f2cd/Snakefile\", line 2, in \u003cmodule\u003e\n File \"/tmp/tmpxxvt0ijh/Reed-CompBio-spras-ae3f2cd/spras/runner.py\", line 2, in \u003cmodule\u003e\n File \"/tmp/tmpxxvt0ijh/Reed-CompBio-spras-ae3f2cd/spras/allpairs.py\", line 4, in \u003cmodule\u003e\n File \"/tmp/tmpxxvt0ijh/Reed-CompBio-spras-ae3f2cd/spras/containers.py\", line 7, in \u003cmodule\u003e\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 11, + "subscribers_count": 7, + "topics": [ + "workkflow", + "snakemake", + "network", + "omics" + ], + "updated_at": 1725994080.0 + }, { "config_readme": "\u003cp\u003ePlease follow the instructions in the template config.yaml that is copied to your deployment via snakedeploy.\u003c/p\u003e\n", "data_format": 2, @@ -53187,28 +53260,6 @@ var data = ], "updated_at": 1725251889.0 }, - { - "config_readme": null, - "data_format": 2, - "description": "Signaling Pathway Reconstruction Analysis Streamliner (SPRAS)", - "formatting": "[DEBUG] \n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\w\u0027\n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\w\u0027\n[DEBUG] In file \"/tmp/tmpxxvt0ijh/Reed-CompBio-spras-ae3f2cd/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "Reed-CompBio/spras", - "latest_release": "0.2.0", - "linting": "/tmp/tmpxxvt0ijh/Reed-CompBio-spras-ae3f2cd/Snakefile:14: SyntaxWarning: invalid escape sequence \u0027\\w\u0027\n params=\"params-\\w+\",\n/tmp/tmpxxvt0ijh/Reed-CompBio-spras-ae3f2cd/Snakefile:15: SyntaxWarning: invalid escape sequence \u0027\\w\u0027\n dataset=\"\\w+\"\nModuleNotFoundError in file /tmp/tmpxxvt0ijh/Reed-CompBio-spras-ae3f2cd/Snakefile, line 2:\nNo module named \u0027docker\u0027\n File \"/tmp/tmpxxvt0ijh/Reed-CompBio-spras-ae3f2cd/Snakefile\", line 2, in \u003cmodule\u003e\n File \"/tmp/tmpxxvt0ijh/Reed-CompBio-spras-ae3f2cd/spras/runner.py\", line 2, in \u003cmodule\u003e\n File \"/tmp/tmpxxvt0ijh/Reed-CompBio-spras-ae3f2cd/spras/allpairs.py\", line 4, in \u003cmodule\u003e\n File \"/tmp/tmpxxvt0ijh/Reed-CompBio-spras-ae3f2cd/spras/containers.py\", line 7, in \u003cmodule\u003e\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 11, - "subscribers_count": 7, - "topics": [ - "workkflow", - "snakemake", - "network", - "omics" - ], - "updated_at": 1725994080.0 - }, { "config_readme": null, "data_format": 2, @@ -53923,6 +53974,58 @@ var data = ], "updated_at": 1620029618.0 }, + { + "config_readme": null, + "data_format": 2, + "description": "MR.PARETO - Modules \u0026 Recipes for Pragmatic Augmentation of Research Efficiency Towards Optimum", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp_5d_5npp/workflow/rules/ATACseq.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmp_5d_5npp/workflow/Snakefile\": Keyword \"input\" at line 38 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[ERROR] In file \"/tmp/tmp_5d_5npp/workflow/Snakefile\": SyntaxError: L43: Unrecognised keyword \u0027rules\u0027 in rule definition\n[DEBUG] In file \"/tmp/tmp_5d_5npp/workflow/Snakefile\": \n[DEBUG] In file \"/tmp/tmp_5d_5npp/workflow/rules/MyData.smk\": Formatted content is different from original\n[INFO] 1 file(s) raised parsing errors \ud83e\udd15\n[INFO] 2 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "epigen/mr.pareto", + "latest_release": null, + "linting": "FileNotFoundError in file /tmp/tmp_5d_5npp/workflow/Snakefile, line 24:\n[Errno 2] No such file or directory: \u0027config/RNAseq/RNAseq_rnaseq_pipeline_config.yaml\u0027\n File \"/tmp/tmp_5d_5npp/workflow/Snakefile\", line 24, in \u003cmodule\u003e\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 15, + "subscribers_count": 6, + "topics": [ + "automation", + "bioinformatics", + "biomedical", + "data-science", + "framework", + "snakemake", + "workflows", + "epigenetics", + "genetics", + "genomics", + "modules", + "next-generation-sequencing", + "python", + "r", + "recipes", + "research", + "science" + ], + "updated_at": 1726071263.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "A metatranscriptomic pipeline optimized for the study of microeukaryotes.", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp7kwvm82s/AlexanderLabWHOI-eukrhythmic-06b7f63/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "AlexanderLabWHOI/eukrhythmic", + "latest_release": "v1.0", + "linting": "FileNotFoundError in file /tmp/tmp7kwvm82s/AlexanderLabWHOI-eukrhythmic-06b7f63/Snakefile, line 15:\n[Errno 2] No such file or directory: \u0027input/metaT_cohen.txt\u0027\n File \"/tmp/tmp7kwvm82s/AlexanderLabWHOI-eukrhythmic-06b7f63/Snakefile\", line 15, in \u003cmodule\u003e\n File \"/tmp/tmp7kwvm82s/AlexanderLabWHOI-eukrhythmic-06b7f63/scripts/importworkspace.py\", line 77, in \u003cmodule\u003e\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/pandas/io/parsers/readers.py\", line 1026, in read_csv\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/pandas/io/parsers/readers.py\", line 620, in _read\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/pandas/io/parsers/readers.py\", line 1620, in __init__\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/pandas/io/parsers/readers.py\", line 1880, in _make_engine\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/pandas/io/common.py\", line 873, in get_handle\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 15, + "subscribers_count": 5, + "topics": [], + "updated_at": 1726022302.0 + }, { "config_readme": null, "data_format": 2, @@ -53973,58 +54076,6 @@ var data = ], "updated_at": 1725181639.0 }, - { - "config_readme": null, - "data_format": 2, - "description": "A metatranscriptomic pipeline optimized for the study of microeukaryotes.", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp7kwvm82s/AlexanderLabWHOI-eukrhythmic-06b7f63/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "AlexanderLabWHOI/eukrhythmic", - "latest_release": "v1.0", - "linting": "FileNotFoundError in file /tmp/tmp7kwvm82s/AlexanderLabWHOI-eukrhythmic-06b7f63/Snakefile, line 15:\n[Errno 2] No such file or directory: \u0027input/metaT_cohen.txt\u0027\n File \"/tmp/tmp7kwvm82s/AlexanderLabWHOI-eukrhythmic-06b7f63/Snakefile\", line 15, in \u003cmodule\u003e\n File \"/tmp/tmp7kwvm82s/AlexanderLabWHOI-eukrhythmic-06b7f63/scripts/importworkspace.py\", line 77, in \u003cmodule\u003e\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/pandas/io/parsers/readers.py\", line 1026, in read_csv\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/pandas/io/parsers/readers.py\", line 620, in _read\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/pandas/io/parsers/readers.py\", line 1620, in __init__\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/pandas/io/parsers/readers.py\", line 1880, in _make_engine\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/pandas/io/common.py\", line 873, in get_handle\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 15, - "subscribers_count": 5, - "topics": [], - "updated_at": 1726022302.0 - }, - { - "config_readme": null, - "data_format": 2, - "description": "MR.PARETO - Modules \u0026 Recipes for Pragmatic Augmentation of Research Efficiency Towards Optimum", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmp9mpkh0fm/workflow/rules/ATACseq.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmp9mpkh0fm/workflow/rules/MyData.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmp9mpkh0fm/workflow/Snakefile\": Keyword \"input\" at line 38 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[ERROR] In file \"/tmp/tmp9mpkh0fm/workflow/Snakefile\": SyntaxError: L43: Unrecognised keyword \u0027rules\u0027 in rule definition\n[INFO] In file \"/tmp/tmp9mpkh0fm/workflow/Snakefile\": 1 file(s) raised parsing errors \ud83e\udd15\n[INFO] In file \"/tmp/tmp9mpkh0fm/workflow/Snakefile\": 2 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "epigen/mr.pareto", - "latest_release": null, - "linting": "Traceback (most recent call last):\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/cli.py\", line 2024, in args_to_api\n any_lint = workflow_api.lint()\n ^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/api.py\", line 337, in _handle_no_dag\n return method(self, *args, **kwargs)\n ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/api.py\", line 354, in lint\n workflow.include(\n\n File \"/home/runner/micromamba-root/envs/snakemake-workflow-catalog/lib/python3.12/site-packages/snakemake/workflow.py\", line 1412, in include\n exec(compile(code, snakefile.get_path_or_uri(), \"exec\"), self.globals)\n\n File \"/tmp/tmp9mpkh0fm/workflow/Snakefile\", line 28, in \u003cmodule\u003e\n print(exc)\n\nFileNotFoundError: [Errno 2] No such file or directory: \u0027config/RNAseq/RNAseq_rnaseq_pipeline_config.yaml\u0027\n\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 15, - "subscribers_count": 6, - "topics": [ - "automation", - "bioinformatics", - "biomedical", - "data-science", - "framework", - "snakemake", - "workflows", - "epigenetics", - "genetics", - "genomics", - "modules", - "next-generation-sequencing", - "python", - "r", - "recipes", - "research", - "science" - ], - "updated_at": 1723646641.0 - }, { "config_readme": null, "data_format": 2, @@ -54441,10 +54492,10 @@ var data = "config_readme": null, "data_format": 2, "description": null, - "formatting": "[DEBUG] \n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\.\u0027\n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\.\u0027\n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\.\u0027\n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\.\u0027\n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\.\u0027\n[WARNING] In file \"/tmp/tmp7o46etrx/pypsa-meets-earth-pypsa-earth-sec-719e417/Snakefile\": Keyword \"output\" at line 129 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmp7o46etrx/pypsa-meets-earth-pypsa-earth-sec-719e417/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "formatting": "[DEBUG] \n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\.\u0027\n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\.\u0027\n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\.\u0027\n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\.\u0027\n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\.\u0027\n[WARNING] In file \"/tmp/tmpqrkr2r5b/pypsa-meets-earth-pypsa-earth-sec-1bcf1e5/Snakefile\": Keyword \"output\" at line 178 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpqrkr2r5b/pypsa-meets-earth-pypsa-earth-sec-1bcf1e5/Snakefile\": Keyword \"output\" at line 962 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] In file \"/tmp/tmpqrkr2r5b/pypsa-meets-earth-pypsa-earth-sec-1bcf1e5/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", "full_name": "pypsa-meets-earth/pypsa-earth-sec", - "latest_release": "0.0.2", - "linting": "SyntaxError in file /tmp/tmp7o46etrx/pypsa-meets-earth-pypsa-earth-sec-719e417/Snakefile, line 37:\nKeyword subworkflow is deprecated. Use module directive instead :\nNone\n", + "latest_release": "0.1.0", + "linting": "SyntaxError in file /tmp/tmpqrkr2r5b/pypsa-meets-earth-pypsa-earth-sec-1bcf1e5/Snakefile, line 57:\nKeyword subworkflow is deprecated. Use module directive instead :\nNone\n", "mandatory_flags": [], "report": false, "software_stack_deployment": null, @@ -54452,7 +54503,7 @@ var data = "stargazers_count": 19, "subscribers_count": 4, "topics": [], - "updated_at": 1722861906.0 + "updated_at": 1726059102.0 }, { "config_readme": null, @@ -54492,7 +54543,7 @@ var data = "config_readme": "\u003cdiv class=\"markdown-heading\"\u003e\u003ch1 class=\"heading-element\"\u003eGeneral settings\u003c/h1\u003e\u003ca id=\"user-content-general-settings\" class=\"anchor\" aria-label=\"Permalink: General settings\" href=\"#general-settings\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eTo configure this workflow, modify \u003ccode\u003econfig/config.yaml\u003c/code\u003e according to your\nneeds, following the explanations provided in the file.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch2 class=\"heading-element\"\u003eSample sheet\u003c/h2\u003e\u003ca id=\"user-content-sample-sheet\" class=\"anchor\" aria-label=\"Permalink: Sample sheet\" href=\"#sample-sheet\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eThe sample sheet contains all samples to be analyzed by UnCoVar.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch2 class=\"heading-element\"\u003eAuto filling\u003c/h2\u003e\u003ca id=\"user-content-auto-filling\" class=\"anchor\" aria-label=\"Permalink: Auto filling\" href=\"#auto-filling\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eUnCoVar offers the possibility to automatically append samples to the sample\nsheet. To load your data into the workflow execute\u003c/p\u003e\n\u003cpre\u003e\u003ccode\u003esnakemake --cores all --use-conda update_sample\n\u003c/code\u003e\u003c/pre\u003e\n\u003cp\u003ewith the root of the UnCoVar as working directory. It is recommended to use\nthe following structure to when adding data automatically:\u003c/p\u003e\n\u003cpre\u003e\u003ccode\u003e\u251c\u2500\u2500 archive\n\u251c\u2500\u2500 incoming\n\u2514\u2500\u2500 snakemake-workflow-sars-cov2\n \u251c\u2500\u2500 data\n \u2514\u2500\u2500 ...\n\u003c/code\u003e\u003c/pre\u003e\n\u003cp\u003eHowever, this structure is not set in stone and can be adjusted via the\n\u003ccode\u003econfig/config.yaml\u003c/code\u003e file under \u003ccode\u003edata-handling\u003c/code\u003e. Only the following path to the\ncorresponding folders, relative to the directory of UnCoVar are needed:\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cstrong\u003eincoming\u003c/strong\u003e: path of incoming data, which is moved to the data directory by\nthe preprocessing script. Defaults to \u003ccode\u003e../incoming/\u003c/code\u003e.\u003c/li\u003e\n\u003cli\u003e\n\u003cstrong\u003edata\u003c/strong\u003e: path to store data within the workflow. defaults to \u003ccode\u003edata/\u003c/code\u003e.\u003c/li\u003e\n\u003cli\u003e\n\u003cstrong\u003earchive\u003c/strong\u003e: path to archive data from the results from the analysis to.\nDefaults to \u003ccode\u003e../archive/\u003c/code\u003e.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eThe incoming directory should contain paired end reads in (compressed) FASTQ\nformat. UnCoVar automatically copies your data into the data directory and moves\nall files from incoming directory to the archive. After the analysis, all results\nare compressed and saved alongside the reads.\u003c/p\u003e\n\u003cp\u003eMoreover, the sample sheet is automatically updated with the new files. Please\nnote, that only the part of the filename before the first \u0027_\u0027 character is used\nas the sample name within the workflow.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch2 class=\"heading-element\"\u003eManual filling\u003c/h2\u003e\u003ca id=\"user-content-manual-filling\" class=\"anchor\" aria-label=\"Permalink: Manual filling\" href=\"#manual-filling\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eOf course, samples to be analyzed can also be added manually to the sample sheet.\nFor each sample, the a new line in \u003ccode\u003econfig/pep/samples.csv\u003c/code\u003e with the following\ncontent has to be defined:\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003cstrong\u003esample_name\u003c/strong\u003e: name or identifier of sample\u003c/li\u003e\n\u003cli\u003e\n\u003cstrong\u003efq1\u003c/strong\u003e: path to read 1 in FASTQ format\u003c/li\u003e\n\u003cli\u003e\n\u003cstrong\u003efq2\u003c/strong\u003e: path to read 2 in FASTQ format\u003c/li\u003e\n\u003cli\u003e\n\u003cstrong\u003edate\u003c/strong\u003e: sampling date of the sample\u003c/li\u003e\n\u003cli\u003e\n\u003cstrong\u003eis_amplicon_data\u003c/strong\u003e: indicates whether the data was generated with a\nshotgun (0) or amplicon (1) sequencing\u003c/li\u003e\n\u003c/ul\u003e\n", "data_format": 2, "description": "Transparent and robust SARS-CoV-2 variant calling and lineage assignment with comprehensive reporting.", - "formatting": "[DEBUG] \n[DEBUG] \n[DEBUG] In file \"/tmp/tmpod2hwe86/IKIM-Essen-uncovar-53b99c4/workflow/rules/common.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] In file \"/tmp/tmpod2hwe86/IKIM-Essen-uncovar-53b99c4/workflow/rules/generate_output.smk\": Formatted content is different from original\n[DEBUG] \n[WARNING] In file \"/tmp/tmpod2hwe86/IKIM-Essen-uncovar-53b99c4/workflow/rules/qc.smk\": Keyword \"input\" at line 38 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpod2hwe86/IKIM-Essen-uncovar-53b99c4/workflow/rules/qc.smk\": Keyword \"input\" at line 62 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] \n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\S\u0027\n[DEBUG] In file \"/tmp/tmpod2hwe86/IKIM-Essen-uncovar-53b99c4/workflow/rules/ref.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n\u003cunknown\u003e:2: SyntaxWarning: invalid escape sequence \u0027\\#\u0027\n[DEBUG] In file \"/tmp/tmpod2hwe86/IKIM-Essen-uncovar-53b99c4/workflow/rules/variant_calling.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] \n[DEBUG] \n[INFO] 4 file(s) would be changed \ud83d\ude2c\n[INFO] 17 file(s) would be left unchanged \ud83c\udf89\n\nsnakefmt version: 0.10.2", + "formatting": "[DEBUG] \n[DEBUG] \n[DEBUG] \n[WARNING] In file \"/tmp/tmpi_dd5cya/IKIM-Essen-uncovar-53b99c4/workflow/rules/qc.smk\": Keyword \"input\" at line 38 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[WARNING] In file \"/tmp/tmpi_dd5cya/IKIM-Essen-uncovar-53b99c4/workflow/rules/qc.smk\": Keyword \"input\" at line 62 has comments under a value.\n\tPEP8 recommends block comments appear before what they describe\n(see https://www.python.org/dev/peps/pep-0008/#id30)\n[DEBUG] \n[DEBUG] \n[DEBUG] \n\u003cunknown\u003e:1: SyntaxWarning: invalid escape sequence \u0027\\S\u0027\n[DEBUG] In file \"/tmp/tmpi_dd5cya/IKIM-Essen-uncovar-53b99c4/workflow/rules/ref.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n\u003cunknown\u003e:2: SyntaxWarning: invalid escape sequence \u0027\\#\u0027\n[DEBUG] In file \"/tmp/tmpi_dd5cya/IKIM-Essen-uncovar-53b99c4/workflow/rules/variant_calling.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpi_dd5cya/IKIM-Essen-uncovar-53b99c4/workflow/rules/generate_output.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] In file \"/tmp/tmpi_dd5cya/IKIM-Essen-uncovar-53b99c4/workflow/rules/common.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] \n[INFO] 4 file(s) would be changed \ud83d\ude2c\n[INFO] 17 file(s) would be left unchanged \ud83c\udf89\n\nsnakefmt version: 0.10.2", "full_name": "IKIM-Essen/uncovar", "latest_release": "v1.0.0", "linting": null, @@ -54509,16 +54560,16 @@ var data = "variant-calling", "lineage-assignment" ], - "updated_at": 1724926685.0 + "updated_at": 1726068377.0 }, { "config_readme": "\u003cdiv class=\"markdown-heading\"\u003e\u003ch1 class=\"heading-element\"\u003eConfiguration\u003c/h1\u003e\u003ca id=\"user-content-configuration\" class=\"anchor\" aria-label=\"Permalink: Configuration\" href=\"#configuration\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eYou need one configuration file to configure the analyses and one annotation file describing the data to run the complete workflow. If in doubt read the comments in the config and/or try the default values. We provide a full example including data and configuration in \u003ccode\u003etest/\u003c/code\u003e as a starting point.\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eproject configuration (\u003ccode\u003econfig/config.yaml\u003c/code\u003e): Different for every project and configures the analyses to be performed.\u003c/li\u003e\n\u003cli\u003esample annotation (annotation): CSV file consisting of four mandatory columns.\n\u003cul\u003e\n\u003cli\u003ename: A unique name of the dataset (tip: keep it short but descriptive).\u003c/li\u003e\n\u003cli\u003edata: Path to the tabular data as comma separated table (CSV).\u003c/li\u003e\n\u003cli\u003emetadata: Path to the metadata as comma separated table (CSV) with the first column being the index/identifier of each observation/sample and every other column metadata for the respective observation (either numeric or categorical, not mixed). \u003cstrong\u003eNo NaN or empty values allowed.\u003c/strong\u003e\n\u003c/li\u003e\n\u003cli\u003esamples_by_features: Boolean indicator if the data matrix is observations/samples (rows) x features (columns): 0==no, 1==yes.\u003c/li\u003e\n\u003c/ul\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n", "data_format": 2, "description": "A general purpose Snakemake workflow to perform unsupervised analyses (dimensionality reduction \u0026 cluster analysis) and visualizations of high-dimensional data.", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpj8z5609o/epigen-unsupervised_analysis-1f3993a/workflow/rules/dimred.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpj8z5609o/epigen-unsupervised_analysis-1f3993a/workflow/rules/visualization.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpj8z5609o/epigen-unsupervised_analysis-1f3993a/workflow/rules/envs_export.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpj8z5609o/epigen-unsupervised_analysis-1f3993a/workflow/rules/clustering.smk\": Formatted content is different from original\n[DEBUG] \n[ERROR] In file \"/tmp/tmpj8z5609o/epigen-unsupervised_analysis-1f3993a/workflow/Snakefile\": InvalidPython: Black error:\n```\nCannot parse: 167:0: EOF in multi-line statement\n```\n\n[DEBUG] In file \"/tmp/tmpj8z5609o/epigen-unsupervised_analysis-1f3993a/workflow/Snakefile\": \n[DEBUG] In file \"/tmp/tmpj8z5609o/epigen-unsupervised_analysis-1f3993a/workflow/rules/cluster_validation.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmpj8z5609o/epigen-unsupervised_analysis-1f3993a/workflow/rules/common.smk\": Formatted content is different from original\n[INFO] 1 file(s) raised parsing errors \ud83e\udd15\n[INFO] 6 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmprxul5lic/epigen-unsupervised_analysis-1f3993a/workflow/rules/visualization.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmprxul5lic/epigen-unsupervised_analysis-1f3993a/workflow/rules/common.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmprxul5lic/epigen-unsupervised_analysis-1f3993a/workflow/rules/cluster_validation.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmprxul5lic/epigen-unsupervised_analysis-1f3993a/workflow/rules/envs_export.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] In file \"/tmp/tmprxul5lic/epigen-unsupervised_analysis-1f3993a/workflow/rules/clustering.smk\": Formatted content is different from original\n[DEBUG] \n[ERROR] In file \"/tmp/tmprxul5lic/epigen-unsupervised_analysis-1f3993a/workflow/Snakefile\": InvalidPython: Black error:\n```\nCannot parse: 167:0: EOF in multi-line statement\n```\n\n[DEBUG] In file \"/tmp/tmprxul5lic/epigen-unsupervised_analysis-1f3993a/workflow/Snakefile\": \n[DEBUG] In file \"/tmp/tmprxul5lic/epigen-unsupervised_analysis-1f3993a/workflow/rules/dimred.smk\": Formatted content is different from original\n[INFO] 1 file(s) raised parsing errors \ud83e\udd15\n[INFO] 6 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", "full_name": "epigen/unsupervised_analysis", "latest_release": "v2.0.0", - "linting": "Lints for rule aggregate_clustering_results (line 101, /tmp/tmpj8z5609o/epigen-unsupervised_analysis-1f3993a/workflow/rules/clustering.smk):\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule annot_export (line 88, /tmp/tmpj8z5609o/epigen-unsupervised_analysis-1f3993a/workflow/rules/envs_export.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\n", + "linting": "Lints for rule aggregate_clustering_results (line 45, /tmp/tmprxul5lic/epigen-unsupervised_analysis-1f3993a/workflow/rules/clustering.smk):\n * Migrate long run directives into scripts or notebooks:\n Long run directives hamper workflow readability. Use the script or\n notebook directive instead. Note that the script or notebook directive\n does not involve boilerplate. Similar to run, you will have direct access\n to params, input, output, and wildcards.Only use the run directive for a\n handful of lines.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#external-scripts\n https://snakemake.readthedocs.io/en/latest/snakefiles/rules.html#jupyter-notebook-integration\n\nLints for rule annot_export (line 43, /tmp/tmprxul5lic/epigen-unsupervised_analysis-1f3993a/workflow/rules/envs_export.smk):\n * Specify a conda environment or container for each rule.:\n This way, the used software for each specific step is documented, and the\n workflow can be executed on any machine without prerequisites.\n Also see:\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#integrated-package-management\n https://snakemake.readthedocs.io/en/latest/snakefiles/deployment.html#running-jobs-in-containers\n\n", "mandatory_flags": null, "report": true, "software_stack_deployment": { @@ -54548,7 +54599,7 @@ var data = "clustree", "leiden-algorithm" ], - "updated_at": 1724828940.0 + "updated_at": 1726047432.0 }, { "config_readme": null, @@ -55514,28 +55565,11 @@ var data = "topics": [], "updated_at": 1724002679.0 }, - { - "config_readme": null, - "data_format": 2, - "description": "CADD scripts release for offline scoring. For more information about CADD, please visit our website", - "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpn5wl80um/kircherlab-CADD-scripts-203ee3b/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", - "full_name": "kircherlab/CADD-scripts", - "latest_release": "v1.7", - "linting": "WorkflowError:\nError validating config file.\nValidationError: \u0027REFERENCEpath\u0027 is a required property\n\nFailed validating \u0027required\u0027 in schema[\u0027allOf\u0027][0][\u0027then\u0027]:\n {\u0027required\u0027: [\u0027REFERENCEpath\u0027, \u0027MMSPLICEpath\u0027]}\n\nOn instance:\n {\u0027Header\u0027: \u0027##CADD GRCh37-v1.7 (c) University of Washington, \u0027\n \u0027Hudson-Alpha Institute for Biotechnology and Berlin \u0027\n \u0027Institute of Health at Charite - Universitaetsmedizin \u0027\n \u0027Berlin 2013-2023. All rights reserved.\u0027,\n \u0027Annotation\u0027: False,\n \u0027EnsemblDB\u0027: 110,\n \u0027ESMmodels\u0027: [\u0027esm1v_t33_650M_UR90S_1\u0027,\n \u0027esm1v_t33_650M_UR90S_2\u0027,\n \u0027esm1v_t33_650M_UR90S_3\u0027,\n \u0027esm1v_t33_650M_UR90S_4\u0027,\n \u0027esm1v_t33_650M_UR90S_5\u0027],\n \u0027ESMbatchsize\u0027: 1}\n", - "mandatory_flags": [], - "report": false, - "software_stack_deployment": null, - "standardized": false, - "stargazers_count": 64, - "subscribers_count": 5, - "topics": [], - "updated_at": 1725458538.0 - }, { "config_readme": "\u003cdiv class=\"markdown-heading\"\u003e\u003ch1 class=\"heading-element\"\u003eGeneral settings\u003c/h1\u003e\u003ca id=\"user-content-general-settings\" class=\"anchor\" aria-label=\"Permalink: General settings\" href=\"#general-settings\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eTo configure this workflow, modify the following files to reflect your dataset and differential expression analysis model:\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003econfig/samples.tsv\u003c/code\u003e: samples sheet with covariates and conditions\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003econfig/units.tsv\u003c/code\u003e: (sequencing) units sheet with raw data paths\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003econfig/config.yaml\u003c/code\u003e: general workflow configuration and differential expression model setup\u003c/li\u003e\n\u003c/ul\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch2 class=\"heading-element\"\u003esamples sheet\u003c/h2\u003e\u003ca id=\"user-content-samples-sheet\" class=\"anchor\" aria-label=\"Permalink: samples sheet\" href=\"#samples-sheet\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eFor each biological sample, add a line to the sample sheet in \u003ccode\u003econfig/samples.tsv\u003c/code\u003e.\nThe column \u003ccode\u003esample\u003c/code\u003e is required and gives the sample name.\nAdditional columns can specify covariates (including batch effects) and conditions.\nThese columns can then be used in the \u003ccode\u003ediffexp: models:\u003c/code\u003e specification section in \u003ccode\u003econfig/config.yaml\u003c/code\u003e (see below)\u003c/p\u003e\n\u003cp\u003eMissing values can be specified by empty columns or by writing \u003ccode\u003eNA\u003c/code\u003e.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch2 class=\"heading-element\"\u003eunits sheet\u003c/h2\u003e\u003ca id=\"user-content-units-sheet\" class=\"anchor\" aria-label=\"Permalink: units sheet\" href=\"#units-sheet\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eFor each sample, add one or more sequencing unit lines (runs, lanes or replicates) to the unit sheet in \u003ccode\u003econfig/units.tsv\u003c/code\u003e.\nFor each unit, provide either of the following:\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003eThe path to two pairead-end FASTQ files in the columns \u003ccode\u003efq1\u003c/code\u003e, \u003ccode\u003efq2\u003c/code\u003e.\u003c/li\u003e\n\u003cli\u003eThe path to a single-end FASTQ file in the column \u003ccode\u003efq1\u003c/code\u003e.\nFor single-end data, you also need to specify \u003ccode\u003efragment_len_mean\u003c/code\u003e and \u003ccode\u003efragment_len_sd\u003c/code\u003e, which should usually be available from your sequencing provider.\u003c/li\u003e\n\u003cli\u003eThe path to a single-end BAM file in the column \u003ccode\u003ebam_single\u003c/code\u003e\n\u003c/li\u003e\n\u003cli\u003eThe path to a paired-end bam BAM file in the column \u003ccode\u003ebam_paired\u003c/code\u003e\n\u003c/li\u003e\n\u003c/ul\u003e\n\u003cp\u003eMissing values can be specified by empty columns or by writing \u003ccode\u003eNA\u003c/code\u003e.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch2 class=\"heading-element\"\u003econfig.yaml\u003c/h2\u003e\u003ca id=\"user-content-configyaml\" class=\"anchor\" aria-label=\"Permalink: config.yaml\" href=\"#configyaml\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eThis file contains the general workflow configuration and the setup for the differential expression analysis performed by sleuth.\nConfigurable options should be explained in the comments above the respective entry or right here in this \u003ccode\u003econfig/README.md\u003c/code\u003e section.\nIf something is unclear, don\u0027t hesitate to \u003ca href=\"https://github.com/snakemake-workflows/rna-seq-kallisto-sleuth/issues/new/choose\"\u003efile an issue in the \u003ccode\u003erna-seq-kallisto-sleuth\u003c/code\u003e GitHub repository\u003c/a\u003e.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch3 class=\"heading-element\"\u003edifferential expression model setup\u003c/h3\u003e\u003ca id=\"user-content-differential-expression-model-setup\" class=\"anchor\" aria-label=\"Permalink: differential expression model setup\" href=\"#differential-expression-model-setup\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eThe core functionality of this workflow is provided by the software \u003ca href=\"https://pachterlab.github.io/sleuth/about\" rel=\"nofollow\"\u003e\u003ccode\u003esleuth\u003c/code\u003e\u003c/a\u003e.\nYou can use it to test for differential expression of genes or transcripts between two or more subgroups of samples.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch4 class=\"heading-element\"\u003emain sleuth model\u003c/h4\u003e\u003ca id=\"user-content-main-sleuth-model\" class=\"anchor\" aria-label=\"Permalink: main sleuth model\" href=\"#main-sleuth-model\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eThe main idea of sleuth\u0027s internal model, is to test a \u003ccode\u003efull:\u003c/code\u003e model (containing (a) variable(s) of interest AND batch effects) against a \u003ccode\u003ereduced:\u003c/code\u003e model (containing ONLY the batch effects).\nSo these are the most important entries to set up under any model that you specify via \u003ccode\u003ediffexp: models:\u003c/code\u003e.\nIf you don\u0027t know any batch effects, the \u003ccode\u003ereduced:\u003c/code\u003e model will have to be \u003ccode\u003e~1\u003c/code\u003e.\nOtherwise it will be the tilde followed by an addition of the names of any columns that contain batch effects, for example: \u003ccode\u003ereduced: ~batch_effect_1 + batch_effect_2\u003c/code\u003e.\nThe full model than additionally includes variables of interest, so fore example: \u003ccode\u003efull: ~variable_of_interest + batch_effect_1 + batch_effect_2\u003c/code\u003e.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch4 class=\"heading-element\"\u003esleuth effect sizes\u003c/h4\u003e\u003ca id=\"user-content-sleuth-effect-sizes\" class=\"anchor\" aria-label=\"Permalink: sleuth effect sizes\" href=\"#sleuth-effect-sizes\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eEffect size estimates are calculated as so-called beta-values by \u003ccode\u003esleuth\u003c/code\u003e.\nFor binary comparisons (your variable of interest has two factor levels), they resemble a log2 fold change.\nTo know which variable of interest to use for the effect size calculation, you need to provide its column name as the \u003ccode\u003eprimary_variable:\u003c/code\u003e.\nAnd for sleuth to know what level of that variable of interest to use as the base level, specify the respective entry as the \u003ccode\u003ebase_level:\u003c/code\u003e.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch3 class=\"heading-element\"\u003epreprocessing \u003ccode\u003eparams\u003c/code\u003e\n\u003c/h3\u003e\u003ca id=\"user-content-preprocessing-params\" class=\"anchor\" aria-label=\"Permalink: preprocessing params\" href=\"#preprocessing-params\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eFor \u003cstrong\u003eadapter trimming\u003c/strong\u003e, \u003ccode\u003ecutadapt\u003c/code\u003e is used, with some defaults for standard Illumina data given in the \u003ccode\u003econfig.yaml\u003c/code\u003e.\nFor more details see the comments in the \u003ccode\u003econfig.yaml\u003c/code\u003e or the \u003ca href=\"https://cutadapt.readthedocs.io\" rel=\"nofollow\"\u003e\u003ccode\u003ecutadapt\u003c/code\u003e documentation\u003c/a\u003e.\nFor parameter suggestions for Lexogen 3\u0027 QuantSeq data, see the section below.\u003c/p\u003e\n\u003cp\u003eFor \u003cstrong\u003etranscript quantification\u003c/strong\u003e, \u003ccode\u003ekallisto\u003c/code\u003e is used.\nFor details regarding its command line arguments, see the \u003ca href=\"https://pachterlab.github.io/kallisto/manual\" rel=\"nofollow\"\u003e\u003ccode\u003ekallisto\u003c/code\u003e documentation\u003c/a\u003e.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch4 class=\"heading-element\"\u003eLexogen 3\u0027 QuantSeq data analysis\u003c/h4\u003e\u003ca id=\"user-content-lexogen-3-quantseq-data-analysis\" class=\"anchor\" aria-label=\"Permalink: Lexogen 3\u0027 QuantSeq data analysis\" href=\"#lexogen-3-quantseq-data-analysis\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eFor Lexogen 3\u0027 QuantSeq data analysis, please set \u003ccode\u003eexperiment: 3-prime-rna-seq: activate: true\u003c/code\u003e in the \u003ccode\u003econfig/config.yaml\u003c/code\u003e file.\nFor more information information on Lexogen QuantSeq 3\u0027 sequencing, see: \u003ca href=\"https://www.lexogen.com/quantseq-3mrna-sequencing/\" rel=\"nofollow\"\u003ehttps://www.lexogen.com/quantseq-3mrna-sequencing/\u003c/a\u003e\nIn addition, for Lexogen 3\u0027 FWD QuantSeq data, we recommend setting the \u003ccode\u003eparams: cutadapt-se:\u003c/code\u003e with:\u003c/p\u003e\n\u003cpre\u003e\u003ccode\u003e adapters: \"-a \u0027r1adapter=AGATCGGAAGAGCACACGTCTGAACTCCAGTCAC;min_overlap=7;max_error_rate=0.005\u0027\"\n extra: \"--minimum-length 33 --nextseq-trim=20 --poly-a\"\n\u003c/code\u003e\u003c/pre\u003e\n\u003cp\u003eThis is an adaptation of the \u003ca href=\"https://faqs.lexogen.com/faq/what-is-the-adapter-sequence-i-need-to-use-for-t-1\" rel=\"nofollow\"\u003eLexogen read preprocessing recommendations for 3\u0027 FWD QuantSeq data\u003c/a\u003e.\nChanges to the recommendations are motivated as follows:\u003c/p\u003e\n\u003cul\u003e\n\u003cli\u003e\n\u003ccode\u003e-m\u003c/code\u003e: We switched to the easier to read \u003ccode\u003e--minimum-length\u003c/code\u003e and apply this minimum length globally. In addition, we increase the minimum length to a default of 33 that makes more sense for kallisto quantification.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003e-O\u003c/code\u003e: Instead of this option, minimum overlap is specified per expression.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003e-a \"polyA=A{20}\"\u003c/code\u003e: We replace this with \u003ccode\u003ecutudapt\u003c/code\u003es dedicated option for \u003ccode\u003e--poly-a\u003c/code\u003e tail removal (which is run after adapter trimming).\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003e-a \"QUALITY=G{20}\"\u003c/code\u003e: We replace this with \u003ccode\u003ecutudapt\u003c/code\u003es dedicated option for the removal artifactual trailing \u003ccode\u003eG\u003c/code\u003es in NextSeq and NovaSeq data: \u003ccode\u003e--nextseq-trim=20\u003c/code\u003e.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003e-n\u003c/code\u003e: With the dedicated \u003ccode\u003ecutadapt\u003c/code\u003e options getting applied in the right order, this option is not needed any more.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003e-a \"r1adapter=A{18}AGATCGGAAGAGCACACGTCTGAACTCCAGTCAC;min_overlap=3;max_error_rate=0.100000\"\u003c/code\u003e: We remove A{18}, as this is handled by \u003ccode\u003e--poly-a\u003c/code\u003e. We increase \u003ccode\u003emin_overlap\u003c/code\u003e to 7 and set the \u003ccode\u003emax_error_rate\u003c/code\u003e to the Illumina error rate of about 0.005, both to avoid spurious adapter matches being removed.\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003e-g \"r1adapter=AGATCGGAAGAGCACACGTCTGAACTCCAGTCAC;min_overlap=20\"\u003c/code\u003e: This is not needed any more, as \u003ccode\u003e-a\u003c/code\u003e option will lead to complete removal of read sequence if adapter is found at the start of the read, see: \u003ca href=\"https://cutadapt.readthedocs.io/en/stable/guide.html#rightmost\" rel=\"nofollow\"\u003ehttps://cutadapt.readthedocs.io/en/stable/guide.html#rightmost\u003c/a\u003e\n\u003c/li\u003e\n\u003cli\u003e\n\u003ccode\u003e--discard-trimmed\u003c/code\u003e: We omit this, as the \u003ccode\u003e-a\u003c/code\u003e with the adapter sequence will lead to complete read sequence removal if adapter is found at start, and the \u003ccode\u003e--minimum-length\u003c/code\u003e will then discard such empty reads.\u003c/li\u003e\n\u003c/ul\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch3 class=\"heading-element\"\u003emeta comparisons\u003c/h3\u003e\u003ca id=\"user-content-meta-comparisons\" class=\"anchor\" aria-label=\"Permalink: meta comparisons\" href=\"#meta-comparisons\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eMeta comparisons allow for comparing two full models against each other.\nThe axes represent the log2-fold changes (beta-scores) for the two models, with each point representing a gene.\nPoints on the diagonal indicate no difference between the comparisons, while deviations from the diagonal suggest differences in gene expression between the treatments.\nFor more details see the comments in the \u003ccode\u003econfig.yaml\u003c/code\u003e.\u003c/p\u003e\n", "data_format": 2, "description": "A Snakemake workflow for differential expression analysis of RNA-seq data with Kallisto and Sleuth.", - "formatting": "[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] In file \"/tmp/tmp8nk45vd_/snakemake-workflows-rna-seq-kallisto-sleuth-787376d/workflow/rules/meta_comparisons.smk\": Formatted content is different from original\n[DEBUG] \n[INFO] 1 file(s) would be changed \ud83d\ude2c\n[INFO] 13 file(s) would be left unchanged \ud83c\udf89\n\nsnakefmt version: 0.10.2", + "formatting": "[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] In file \"/tmp/tmpsoaw3fnw/snakemake-workflows-rna-seq-kallisto-sleuth-787376d/workflow/rules/meta_comparisons.smk\": Formatted content is different from original\n[DEBUG] \n[DEBUG] \n[DEBUG] \n[DEBUG] \n[INFO] 1 file(s) would be changed \ud83d\ude2c\n[INFO] 13 file(s) would be left unchanged \ud83c\udf89\n\nsnakefmt version: 0.10.2", "full_name": "snakemake-workflows/rna-seq-kallisto-sleuth", "latest_release": "v2.7.1", "linting": null, @@ -55556,7 +55590,24 @@ var data = "sciworkflows", "reproducibility" ], - "updated_at": 1724920378.0 + "updated_at": 1726071663.0 + }, + { + "config_readme": null, + "data_format": 2, + "description": "CADD scripts release for offline scoring. For more information about CADD, please visit our website", + "formatting": "[DEBUG] \n[DEBUG] In file \"/tmp/tmpn5wl80um/kircherlab-CADD-scripts-203ee3b/Snakefile\": Formatted content is different from original\n[INFO] 1 file(s) would be changed \ud83d\ude2c\n\nsnakefmt version: 0.10.2", + "full_name": "kircherlab/CADD-scripts", + "latest_release": "v1.7", + "linting": "WorkflowError:\nError validating config file.\nValidationError: \u0027REFERENCEpath\u0027 is a required property\n\nFailed validating \u0027required\u0027 in schema[\u0027allOf\u0027][0][\u0027then\u0027]:\n {\u0027required\u0027: [\u0027REFERENCEpath\u0027, \u0027MMSPLICEpath\u0027]}\n\nOn instance:\n {\u0027Header\u0027: \u0027##CADD GRCh37-v1.7 (c) University of Washington, \u0027\n \u0027Hudson-Alpha Institute for Biotechnology and Berlin \u0027\n \u0027Institute of Health at Charite - Universitaetsmedizin \u0027\n \u0027Berlin 2013-2023. All rights reserved.\u0027,\n \u0027Annotation\u0027: False,\n \u0027EnsemblDB\u0027: 110,\n \u0027ESMmodels\u0027: [\u0027esm1v_t33_650M_UR90S_1\u0027,\n \u0027esm1v_t33_650M_UR90S_2\u0027,\n \u0027esm1v_t33_650M_UR90S_3\u0027,\n \u0027esm1v_t33_650M_UR90S_4\u0027,\n \u0027esm1v_t33_650M_UR90S_5\u0027],\n \u0027ESMbatchsize\u0027: 1}\n", + "mandatory_flags": [], + "report": false, + "software_stack_deployment": null, + "standardized": false, + "stargazers_count": 64, + "subscribers_count": 5, + "topics": [], + "updated_at": 1725458538.0 }, { "config_readme": null, @@ -56150,12 +56201,12 @@ var data = "formatting": null, "full_name": "PyPSA/pypsa-eur", "latest_release": "v0.12.0", - "linting": "ModuleNotFoundError in file /tmp/tmpy7hs_fzc/PyPSA-pypsa-eur-0a1d58f/Snakefile, line 13:\nNo module named \u0027tqdm\u0027\n File \"/tmp/tmpy7hs_fzc/PyPSA-pypsa-eur-0a1d58f/Snakefile\", line 13, in \u003cmodule\u003e\n File \"/tmp/tmpy7hs_fzc/PyPSA-pypsa-eur-0a1d58f/scripts/_helpers.py\", line 23, in \u003cmodule\u003e\n", + "linting": "ModuleNotFoundError in file /tmp/tmp369rtxhm/PyPSA-pypsa-eur-0a1d58f/Snakefile, line 13:\nNo module named \u0027tqdm\u0027\n File \"/tmp/tmp369rtxhm/PyPSA-pypsa-eur-0a1d58f/Snakefile\", line 13, in \u003cmodule\u003e\n File \"/tmp/tmp369rtxhm/PyPSA-pypsa-eur-0a1d58f/scripts/_helpers.py\", line 23, in \u003cmodule\u003e\n", "mandatory_flags": [], "report": false, "software_stack_deployment": null, "standardized": false, - "stargazers_count": 318, + "stargazers_count": 320, "subscribers_count": 16, "topics": [ "snakemake", @@ -56167,7 +56218,7 @@ var data = "energy-system-model", "energy-systems" ], - "updated_at": 1725998076.0 + "updated_at": 1726116265.0 }, { "config_readme": "\u003cdiv class=\"markdown-heading\"\u003e\u003ch1 class=\"heading-element\"\u003eGeneral configuration\u003c/h1\u003e\u003ca id=\"user-content-general-configuration\" class=\"anchor\" aria-label=\"Permalink: General configuration\" href=\"#general-configuration\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eTo configure this workflow, modify \u003ccode\u003econfig/config.yaml\u003c/code\u003e according to your needs, following the explanations provided in the file.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch2 class=\"heading-element\"\u003e\n\u003ccode\u003eDESeq2\u003c/code\u003e differential expression analysis configuration\u003c/h2\u003e\u003ca id=\"user-content-deseq2-differential-expression-analysis-configuration\" class=\"anchor\" aria-label=\"Permalink: DESeq2 differential expression analysis configuration\" href=\"#deseq2-differential-expression-analysis-configuration\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eTo successfully run the differential expression analysis, you will need to tell DESeq2 which sample annotations to use (annotations are columns in the \u003ccode\u003esamples.tsv\u003c/code\u003e file described below).\nThis is done in the \u003ccode\u003econfig.yaml\u003c/code\u003e file with the entries under \u003ccode\u003ediffexp:\u003c/code\u003e.\nThe comments for the entries should give all the necessary infos and linkouts.\nBut if in doubt, please also consult the \u003ca href=\"https://www.bioconductor.org/packages/devel/bioc/vignettes/DESeq2/inst/doc/DESeq2.html\" rel=\"nofollow\"\u003e\u003ccode\u003eDESeq2\u003c/code\u003e manual\u003c/a\u003e.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch1 class=\"heading-element\"\u003eSample and unit setup\u003c/h1\u003e\u003ca id=\"user-content-sample-and-unit-setup\" class=\"anchor\" aria-label=\"Permalink: Sample and unit setup\" href=\"#sample-and-unit-setup\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eThe sample and unit setup is specified via tab-separated tabular files (\u003ccode\u003e.tsv\u003c/code\u003e).\nMissing values can be specified by empty columns or by writing \u003ccode\u003eNA\u003c/code\u003e.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch2 class=\"heading-element\"\u003esample sheet\u003c/h2\u003e\u003ca id=\"user-content-sample-sheet\" class=\"anchor\" aria-label=\"Permalink: sample sheet\" href=\"#sample-sheet\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eThe default sample sheet is \u003ccode\u003econfig/samples.tsv\u003c/code\u003e (as configured in \u003ccode\u003econfig/config.yaml\u003c/code\u003e).\nEach sample refers to an actual physical sample, and replicates (both biological and technical) may be specified as separate samples.\nFor each sample, you will always have to specify a \u003ccode\u003esample_name\u003c/code\u003e.\nIn addition, all \u003ccode\u003evariables_of_interest\u003c/code\u003e and \u003ccode\u003ebatch_effects\u003c/code\u003e specified in the \u003ccode\u003econfig/config.yaml\u003c/code\u003e under the \u003ccode\u003ediffexp:\u003c/code\u003e entry, will have to have corresponding columns in the \u003ccode\u003econfig/samples.tsv\u003c/code\u003e.\nFinally, the sample sheet can contain any number of additional columns.\nSo if in doubt about whether you might at some point need some metadata you already have at hand, just put it into the sample sheet already---your future self will thank you.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch2 class=\"heading-element\"\u003eunit sheet\u003c/h2\u003e\u003ca id=\"user-content-unit-sheet\" class=\"anchor\" aria-label=\"Permalink: unit sheet\" href=\"#unit-sheet\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eThe default unit sheet is \u003ccode\u003econfig/units.tsv\u003c/code\u003e (as configured in \u003ccode\u003econfig/config.yaml\u003c/code\u003e).\nFor each sample, add one or more sequencing units (for example if you have several runs or lanes per sample).\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch3 class=\"heading-element\"\u003e\n\u003ccode\u003e.fastq\u003c/code\u003e file source\u003c/h3\u003e\u003ca id=\"user-content-fastq-file-source\" class=\"anchor\" aria-label=\"Permalink: .fastq file source\" href=\"#fastq-file-source\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eFor each unit, you will have to define a source for your \u003ccode\u003e.fastq\u003c/code\u003e files.\nThis can be done via the columns \u003ccode\u003efq1\u003c/code\u003e, \u003ccode\u003efq2\u003c/code\u003e and \u003ccode\u003esra\u003c/code\u003e, with either of:\u003c/p\u003e\n\u003col\u003e\n\u003cli\u003eA single \u003ccode\u003e.fastq\u003c/code\u003e file for single-end reads (\u003ccode\u003efq1\u003c/code\u003e column only; \u003ccode\u003efq2\u003c/code\u003e and \u003ccode\u003esra\u003c/code\u003e columns present, but empty).\nThe entry can be any path on your system, but we suggest something like a \u003ccode\u003eraw/\u003c/code\u003e data directory within your analysis directory.\u003c/li\u003e\n\u003cli\u003eTwo \u003ccode\u003e.fastq\u003c/code\u003e files for paired-end reads (columns \u003ccode\u003efq1\u003c/code\u003e and \u003ccode\u003efq2\u003c/code\u003e; column \u003ccode\u003esra\u003c/code\u003e present, but empty).\nAs for the \u003ccode\u003efq1\u003c/code\u003e column, the \u003ccode\u003efq2\u003c/code\u003e column can also point to anywhere on your system.\u003c/li\u003e\n\u003cli\u003eA sequence read archive (SRA) accession number (\u003ccode\u003esra\u003c/code\u003e column only; \u003ccode\u003efq1\u003c/code\u003e and \u003ccode\u003efq2\u003c/code\u003e columns present, but empty).\nThe workflow will automatically download the corresponding \u003ccode\u003e.fastq\u003c/code\u003e data (currently assumed to be paired-end).\nThe accession numbers usually start with SRR or ERR and you can find accession numbers for studies of interest with the \u003ca href=\"https://trace.ncbi.nlm.nih.gov/Traces/study/\" rel=\"nofollow\"\u003eSRA Run Selector\u003c/a\u003e.\nIf both local files and an SRA accession are specified for the same unit, the local files will be used.\u003c/li\u003e\n\u003c/ol\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch3 class=\"heading-element\"\u003eadapter trimming\u003c/h3\u003e\u003ca id=\"user-content-adapter-trimming\" class=\"anchor\" aria-label=\"Permalink: adapter trimming\" href=\"#adapter-trimming\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eIf you set \u003ccode\u003etrimming: activate:\u003c/code\u003e in the \u003ccode\u003econfig/config.yaml\u003c/code\u003e to \u003ccode\u003eTrue\u003c/code\u003e, you will have to provide at least one \u003ccode\u003ecutadapt\u003c/code\u003e adapter argument for each unit in the \u003ccode\u003eadapters\u003c/code\u003e column of the \u003ccode\u003eunits.tsv\u003c/code\u003e file.\nYou will need to find out the adapters used in the sequencing protocol that generated a unit: from your sequencing provider, or for published data from the study\u0027s metadata (or its authors).\nThen, enter the adapter sequences into the \u003ccode\u003eadapters\u003c/code\u003e column of that unit, preceded by the \u003ca href=\"https://cutadapt.readthedocs.io/en/stable/guide.html#adapter-types\" rel=\"nofollow\"\u003ecorrect \u003ccode\u003ecutadapt\u003c/code\u003e adapter argument\u003c/a\u003e.\u003c/p\u003e\n\u003cdiv class=\"markdown-heading\"\u003e\u003ch3 class=\"heading-element\"\u003estrandedness of library preparation protocol\u003c/h3\u003e\u003ca id=\"user-content-strandedness-of-library-preparation-protocol\" class=\"anchor\" aria-label=\"Permalink: strandedness of library preparation protocol\" href=\"#strandedness-of-library-preparation-protocol\"\u003e\u003cspan aria-hidden=\"true\" class=\"octicon octicon-link\"\u003e\u003c/span\u003e\u003c/a\u003e\u003c/div\u003e\n\u003cp\u003eTo get the correct \u003ccode\u003egeneCounts\u003c/code\u003e from \u003ccode\u003eSTAR\u003c/code\u003e output, you can provide information on the strandedness of the library preparation protocol used for a unit.\n\u003ccode\u003eSTAR\u003c/code\u003e can produce counts for unstranded (\u003ccode\u003enone\u003c/code\u003e - this is the default), forward oriented (\u003ccode\u003eyes\u003c/code\u003e) and reverse oriented (\u003ccode\u003ereverse\u003c/code\u003e) protocols.\u003cbr\u003e\nEnter the respective value into a \u003ccode\u003estrandedness\u003c/code\u003e column in the \u003ccode\u003eunits.tsv\u003c/code\u003e file.\u003c/p\u003e\n", diff --git a/skips.json b/skips.json index 74fbb3b..772c756 100644 --- a/skips.json +++ b/skips.json @@ -1,4 +1,208 @@ [ + { + "full_name": "marbl/verkko", + "updated_at": 1726106396.0 + }, + { + "full_name": "r0xsh/my-awesome-stars", + "updated_at": 1726106263.0 + }, + { + "full_name": "saforem2/awesome-stars", + "updated_at": 1726106121.0 + }, + { + "full_name": "tycloud97/awesome-stars", + "updated_at": 1726106038.0 + }, + { + "full_name": "sunlei/awesome-stars", + "updated_at": 1726102301.0 + }, + { + "full_name": "SaschaLucius/my-awesome-stars", + "updated_at": 1726102080.0 + }, + { + "full_name": "nextstrain/forecasts-ncov", + "updated_at": 1726101510.0 + }, + { + "full_name": "Doudna-lab/snakedali", + "updated_at": 1726101420.0 + }, + { + "full_name": "squisquater/KitFoxGBS-PipelineTest20240910", + "updated_at": 1726101317.0 + }, + { + "full_name": "jbloomlab/seqneut-pipeline", + "updated_at": 1726100217.0 + }, + { + "full_name": "resheku/my-awesome", + "updated_at": 1726099823.0 + }, + { + "full_name": "PacificBiosciences/pb-metagenomics-tools", + "updated_at": 1726096933.0 + }, + { + "full_name": "brss12/APAtizer", + "updated_at": 1726093726.0 + }, + { + "full_name": "DarrianTalamantes/Mapping_and_QTL", + "updated_at": 1726091103.0 + }, + { + "full_name": "oxsecurity/megalinter", + "updated_at": 1726089754.0 + }, + { + "full_name": "EthicalML/awesome-production-machine-learning", + "updated_at": 1726111162.0 + }, + { + "full_name": "hchulkim/replication-template", + "updated_at": 1726081836.0 + }, + { + "full_name": "super-linter/super-linter", + "updated_at": 1726093106.0 + }, + { + "full_name": "jacksonh1/slim_conservation_scoring", + "updated_at": 1726075496.0 + }, + { + "full_name": "Sebastien-Raguideau/Metahood", + "updated_at": 1726075044.0 + }, + { + "full_name": "KaylBing/FindViralStrains", + "updated_at": 1726074734.0 + }, + { + "full_name": "kennedyopokuasare/Azure_datascience", + "updated_at": 1726074564.0 + }, + { + "full_name": "jbloomlab/Herpesvirus-Glycoprotein-Analysis", + "updated_at": 1726073601.0 + }, + { + "full_name": "CCBR/actions", + "updated_at": 1726073324.0 + }, + { + "full_name": "ResearchObject/ro-crate-py", + "updated_at": 1726070801.0 + }, + { + "full_name": "ValeriaDifilippo/NAFuse", + "updated_at": 1726071161.0 + }, + { + "full_name": "calliope-project/ec_modules", + "updated_at": 1726070205.0 + }, + { + "full_name": "jacksonh1/slim_conservation_orthogroup_generation", + "updated_at": 1726068076.0 + }, + { + "full_name": "afids/autoafids_prep", + "updated_at": 1726065830.0 + }, + { + "full_name": "kraemer-lab/GRAPEVNE", + "updated_at": 1726065712.0 + }, + { + "full_name": "rainoffallingstar/BeaverdownContainers", + "updated_at": 1726065574.0 + }, + { + "full_name": "gagneurlab/absplice", + "updated_at": 1726064712.0 + }, + { + "full_name": "jinghuazhao/pQTLtools", + "updated_at": 1726064103.0 + }, + { + "full_name": "santaci/SimOutbreakSelection", + "updated_at": 1726061732.0 + }, + { + "full_name": "epigen/cemm.slurm.sm", + "updated_at": 1726060133.0 + }, + { + "full_name": "Louis-XIV-bis/variant_calling_scere", + "updated_at": 1726057255.0 + }, + { + "full_name": "PyPSA/pypsa-ariadne", + "updated_at": 1726044852.0 + }, + { + "full_name": "ZYX593889540/BsClass", + "updated_at": 1726043555.0 + }, + { + "full_name": "jimbrig/jimsghstars", + "updated_at": 1726109215.0 + }, + { + "full_name": "snakemake/snakemake-hpc-teaching-material", + "updated_at": 1726040344.0 + }, + { + "full_name": "haydeeartaza/RareCNVsAnalysis", + "updated_at": 1726039959.0 + }, + { + "full_name": "Wytamma/snk", + "updated_at": 1726034878.0 + }, + { + "full_name": "Aurametrix/Alg", + "updated_at": 1726016290.0 + }, + { + "full_name": "tbrunetti/library", + "updated_at": 1726016237.0 + }, + { + "full_name": "tycheleturner/ElGenomaPequeno", + "updated_at": 1726012189.0 + }, + { + "full_name": "Schroeder-Group/Parasites-project", + "updated_at": 1726068275.0 + }, + { + "full_name": "ohsu-comp-bio/awesome-tes", + "updated_at": 1725994546.0 + }, + { + "full_name": "maxfieldk/TE-Seq", + "updated_at": 1725994314.0 + }, + { + "full_name": "ralatsdc/nsforest-docker", + "updated_at": 1725988646.0 + }, + { + "full_name": "RTXteam/RTX-KG2", + "updated_at": 1725983524.0 + }, + { + "full_name": "NCI-CGR/MineSV", + "updated_at": 1725979678.0 + }, { "full_name": "gabefoley/asr_curation", "updated_at": 1721110228.0 @@ -331,10 +535,6 @@ "full_name": "nextstrain/rsv", "updated_at": 1721334490.0 }, - { - "full_name": "nextstrain/forecasts-ncov", - "updated_at": 1721334256.0 - }, { "full_name": "maxwendler/masters_thesis", "updated_at": 1721332373.0 @@ -1075,10 +1275,6 @@ "full_name": "Moha-cm/AutoRNASeq", "updated_at": 1724857295.0 }, - { - "full_name": "PacificBiosciences/pb-metagenomics-tools", - "updated_at": 1724449242.0 - }, { "full_name": "ICESAT-2HackWeek/GroundingLine-Greenland", "updated_at": 1724443992.0 @@ -1315,14 +1511,6 @@ "full_name": "dadrasarmin/rnaseq_quantification_kallisto_pipeline", "updated_at": 1724784689.0 }, - { - "full_name": "DarrianTalamantes/Mapping_and_QTL", - "updated_at": 1724781163.0 - }, - { - "full_name": "afids/autoafids_prep", - "updated_at": 1724778175.0 - }, { "full_name": "sylvainschmitt/climfor", "updated_at": 1724772527.0 @@ -1351,10 +1539,6 @@ "full_name": "Diako-Lab/SRAscraper", "updated_at": 1724726805.0 }, - { - "full_name": "Doudna-lab/snakedali", - "updated_at": 1724749955.0 - }, { "full_name": "swajid/review_and_read_ahead", "updated_at": 1724701536.0 @@ -1507,10 +1691,6 @@ "full_name": "rasilab/nugent_2024", "updated_at": 1725127259.0 }, - { - "full_name": "jbloomlab/seqneut-pipeline", - "updated_at": 1725055631.0 - }, { "full_name": "collaborativebioinformatics/starfishDiscovery", "updated_at": 1725328127.0 @@ -1627,10 +1807,6 @@ "full_name": "GABallena/Project4", "updated_at": 1725747736.0 }, - { - "full_name": "brss12/APAtizer", - "updated_at": 1725745214.0 - }, { "full_name": "rraadd88/roux", "updated_at": 1725739970.0 @@ -1727,10 +1903,6 @@ "full_name": "energyLS/pypsa-nz", "updated_at": 1725531058.0 }, - { - "full_name": "Louis-XIV-bis/variant_calling_scere", - "updated_at": 1725526129.0 - }, { "full_name": "gi-bielefeld/corer", "updated_at": 1725522897.0 @@ -1775,122 +1947,10 @@ "full_name": "Robin1em/DTI_Processing_Snakemake", "updated_at": 1725457100.0 }, - { - "full_name": "oxsecurity/megalinter", - "updated_at": 1726025840.0 - }, - { - "full_name": "rainoffallingstar/BeaverdownContainers", - "updated_at": 1726023678.0 - }, - { - "full_name": "ZYX593889540/BsClass", - "updated_at": 1726023173.0 - }, - { - "full_name": "r0xsh/my-awesome-stars", - "updated_at": 1726019843.0 - }, - { - "full_name": "saforem2/awesome-stars", - "updated_at": 1726019692.0 - }, - { - "full_name": "tycloud97/awesome-stars", - "updated_at": 1726019618.0 - }, - { - "full_name": "Aurametrix/Alg", - "updated_at": 1726016290.0 - }, - { - "full_name": "tbrunetti/library", - "updated_at": 1726016237.0 - }, - { - "full_name": "sunlei/awesome-stars", - "updated_at": 1726015877.0 - }, - { - "full_name": "SaschaLucius/my-awesome-stars", - "updated_at": 1726015676.0 - }, - { - "full_name": "resheku/my-awesome", - "updated_at": 1726013402.0 - }, - { - "full_name": "tycheleturner/ElGenomaPequeno", - "updated_at": 1726012189.0 - }, - { - "full_name": "jacksonh1/slim_conservation_orthogroup_generation", - "updated_at": 1726008627.0 - }, - { - "full_name": "marbl/verkko", - "updated_at": 1726003533.0 - }, - { - "full_name": "jacksonh1/slim_conservation_scoring", - "updated_at": 1726002679.0 - }, - { - "full_name": "CCBR/actions", - "updated_at": 1725997456.0 - }, - { - "full_name": "super-linter/super-linter", - "updated_at": 1726014841.0 - }, - { - "full_name": "snakemake/snakemake-hpc-teaching-material", - "updated_at": 1725995191.0 - }, - { - "full_name": "ohsu-comp-bio/awesome-tes", - "updated_at": 1725994546.0 - }, - { - "full_name": "maxfieldk/TE-Seq", - "updated_at": 1725994314.0 - }, - { - "full_name": "kennedyopokuasare/Azure_datascience", - "updated_at": 1725989443.0 - }, - { - "full_name": "ralatsdc/nsforest-docker", - "updated_at": 1725988646.0 - }, - { - "full_name": "RTXteam/RTX-KG2", - "updated_at": 1725983524.0 - }, - { - "full_name": "jinghuazhao/pQTLtools", - "updated_at": 1725983048.0 - }, - { - "full_name": "PyPSA/pypsa-ariadne", - "updated_at": 1725982404.0 - }, - { - "full_name": "haydeeartaza/RareCNVsAnalysis", - "updated_at": 1725980658.0 - }, - { - "full_name": "NCI-CGR/MineSV", - "updated_at": 1725979678.0 - }, { "full_name": "martibosch/cookiecutter-data-snake", "updated_at": 1725976894.0 }, - { - "full_name": "calliope-project/ec_modules", - "updated_at": 1725975827.0 - }, { "full_name": "BradleyH017/atlassing", "updated_at": 1725974025.0 @@ -1907,10 +1967,6 @@ "full_name": "alcrene/smttask", "updated_at": 1725970784.0 }, - { - "full_name": "ResearchObject/ro-crate-py", - "updated_at": 1725970767.0 - }, { "full_name": "pditommaso/awesome-pipeline", "updated_at": 1725986868.0 @@ -1919,10 +1975,6 @@ "full_name": "Hutton-Potato-Genetics/nfHISS", "updated_at": 1725956852.0 }, - { - "full_name": "jimbrig/jimsghstars", - "updated_at": 1725956104.0 - }, { "full_name": "RasmussenLab/vamb", "updated_at": 1725986056.0 @@ -1931,10 +1983,6 @@ "full_name": "yenon118/SQLmireba", "updated_at": 1725935400.0 }, - { - "full_name": "hchulkim/replication-template", - "updated_at": 1725934126.0 - }, { "full_name": "Vimlendu27/super-linter", "updated_at": 1725919645.0 @@ -1947,14 +1995,6 @@ "full_name": "pelayovic/K-CHOPO", "updated_at": 1725902943.0 }, - { - "full_name": "EthicalML/awesome-production-machine-learning", - "updated_at": 1726033587.0 - }, - { - "full_name": "KaylBing/FindViralStrains", - "updated_at": 1725900911.0 - }, { "full_name": "matinnuhamunada/trycycler_snakemake_wrapper", "updated_at": 1725899450.0 @@ -1967,10 +2007,6 @@ "full_name": "unipept/peptonizer", "updated_at": 1725886895.0 }, - { - "full_name": "epigen/cemm.slurm.sm", - "updated_at": 1725886534.0 - }, { "full_name": "shahhet195/mirror-action-github-super-linter", "updated_at": 1725881678.0 @@ -2019,14 +2055,6 @@ "full_name": "BjornFJohansson/pydna", "updated_at": 1724048327.0 }, - { - "full_name": "kraemer-lab/GRAPEVNE", - "updated_at": 1724323350.0 - }, - { - "full_name": "Wytamma/snk", - "updated_at": 1725510522.0 - }, { "full_name": "CIML-bioinformatic/BMMlab_Prostate-CD45Pos-PTENKO", "updated_at": 1724308136.0 @@ -2427,18 +2455,10 @@ "full_name": "BEFH/snakemake-executor-plugin-lsf", "updated_at": 1719308025.0 }, - { - "full_name": "Sebastien-Raguideau/Metahood", - "updated_at": 1718206498.0 - }, { "full_name": "dib-lab/sourmash_plugin_pangenomics", "updated_at": 1718202091.0 }, - { - "full_name": "santaci/SimOutbreakSelection", - "updated_at": 1719500588.0 - }, { "full_name": "nd-nehadaga/ChIP-seq_analysis", "updated_at": 1718200291.0 @@ -2699,10 +2719,6 @@ "full_name": "CompGenomeLab/xr-ds-seq-snakemake", "updated_at": 1719482308.0 }, - { - "full_name": "gagneurlab/absplice", - "updated_at": 1717326831.0 - }, { "full_name": "epigen/mr.pareto", "updated_at": 1720614886.0