Merge pull request #1111 from taylorpaisie/tkp-hostile

Adding hostile v1.1.0

.github/workflows/update_freyja.yml

@@ -86,4 +86,4 @@ jobs:
             quay.io/staphb/freyja:latest
 
       - name: Image digest
-        run: echo ${{ steps.docker_build_user_defined_tag.outputs.digest }}
+        run: echo ${{ steps.docker_build_user_defined_tag.outputs.digest }}

Program_Licenses.md

@@ -73,6 +73,7 @@ The licenses of the open-source software that is contained in these Docker image
 | HeatCluster | GPL-3.0 | https://github.com/DrB-S/heatcluster/blob/main/LICENSE |
 | Hmmer | BSD-3 | http://eddylab.org/software/hmmer/Userguide.pdf |
 | homopolish | GNU GPLv3 | https://github.com/ythuang0522/homopolish/blob/master/LICENSE |
+| hostile | MIT | https://github.com/bede/hostile?tab=MIT-1-ov-file#readme |
 | htslib | MIT | https://github.com/samtools/htslib/blob/develop/LICENSE |
 | Integron Finder | GNU GPLv3 | https://github.com/gem-pasteur/Integron_Finder/blob/master/COPYING |
 | iqtree | GNU GPLv2 | https://github.com/Cibiv/IQ-TREE/blob/master/LICENSE |

README.md

@@ -182,6 +182,7 @@ To learn more about the docker pull rate limits and the open source software pro
 | [heatcluster](https://hub.docker.com/r/staphb/heatcluster) <br/> [![docker pulls](https://badgen.net/docker/pulls/staphb/heatcluster)](https://hub.docker.com/r/staphb/heatcluster) | <ul><li>[1.0.2c](./heatcluster/1.0.2c/)</li></ul> | https://github.com/DrB-S/heatcluster/tree/main |
 | [hmmer](https://hub.docker.com/r/staphb/hmmer) <br/> [![docker pulls](https://badgen.net/docker/pulls/staphb/hmmer)](https://hub.docker.com/r/staphb/hmmer) | <ul><li>[3.3](hmmer/3.3/)</li><li>[3.3.2](hmmer/3.3.2/)</li><li>[3.4](./hmmer/3.4/)</li></ul> | http://hmmer.org/ |
 | [homopolish](https://hub.docker.com/r/staphb/homopolish) <br/> [![docker pulls](https://badgen.net/docker/pulls/staphb/homopolish)](https://hub.docker.com/r/staphb/homopolish) | <ul><li>0.4.1</li></ul> | https://github.com/ythuang0522/homopolish/ |
+| [hostile](https://github.com/bede/hostile) <br/> [![docker pulls](https://badgen.net/docker/pulls/staphb/hostile)](https://hub.docker.com/r/staphb/hostile | <ul><li>[1.1.0](./hostile/1.1.0)</li></ul> | https://github.com/bede/hostile |
 | [htslib](https://hub.docker.com/r/staphb/htslib) <br/> [![docker pulls](https://badgen.net/docker/pulls/staphb/htslib)](https://hub.docker.com/r/staphb/htslib) | <ul><li>[1.14](./htslib/1.14)</li><li>[1.15](./htslib/1.15)</li><li>[1.16](./htslib/1.16)</li><li>[1.17](./htslib/1.17)</li><li>[1.18](./htslib/1.18/)</li><li>[1.19](./htslib/1.19/)</li><li>[1.20](./htslib/1.20/)</li><li>[1.20.c](./htslib/1.20.c/)</li><li>[1.21](./htslib/1.21/)</li></ul> | https://www.htslib.org/ |
 | [Integron Finder](https://hub.docker.com/r/staphb/integron_finder/) <br/> [![docker pulls](https://badgen.net/docker/pulls/staphb/integron_finder)](https://hub.docker.com/r/staphb/integron_finder) | <ul><li>[2.0.5](./integron_finder/2.0.5/)</li></ul> | https://github.com/gem-pasteur/Integron_Finder |
 | [iqtree](https://hub.docker.com/r/staphb/iqtree/) <br/> [![docker pulls](https://badgen.net/docker/pulls/staphb/iqtree)](https://hub.docker.com/r/staphb/iqtree) | <ul><li>1.6.7</li></ul> | http://www.iqtree.org/ |

hostile/1.1.0/Dockerfile

@@ -0,0 +1,74 @@
+# Set global variables
+ARG HOSTILE_VER="1.1.0"
+ARG BOWTIE2_VER="2.5.1"
+
+# Stage 1: Build Dockerfile
+FROM ubuntu:focal AS builder
+ARG HOSTILE_VER
+ARG BOWTIE2_VER
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    software-properties-common && \
+    add-apt-repository ppa:deadsnakes/ppa && \
+    apt-get update && apt-get install -y --no-install-recommends \
+    python3.10 python3.10-distutils build-essential \
+    wget unzip samtools minimap2 bedtools gawk ca-certificates && \
+    apt-get clean && rm -rf /var/lib/apt/lists/*
+
+# Install pip for Python 3.10
+RUN wget https://bootstrap.pypa.io/get-pip.py && python3.10 get-pip.py && \
+    rm get-pip.py && pip install --no-cache-dir setuptools pytest && \
+    ln -sf /usr/bin/python3.10 /usr/bin/python3 && \
+    ln -sf /usr/local/bin/pip /usr/bin/pip
+
+# Install bowtie2 and ensure all executables are accessible
+RUN wget https://github.com/BenLangmead/bowtie2/releases/download/v${BOWTIE2_VER}/bowtie2-${BOWTIE2_VER}-linux-x86_64.zip && \
+    unzip bowtie2-${BOWTIE2_VER}-linux-x86_64.zip && \
+    mv bowtie2-${BOWTIE2_VER}-linux-x86_64 /usr/local/bowtie2 && \
+    ln -s /usr/local/bowtie2/bowtie2* /usr/bin/ && \
+    rm -f bowtie2-${BOWTIE2_VER}-linux-x86_64.zip
+
+# Install hostile
+RUN wget https://github.com/bede/hostile/archive/refs/tags/${HOSTILE_VER}.tar.gz && \
+    tar -xzvf ${HOSTILE_VER}.tar.gz && cd hostile-${HOSTILE_VER} && \
+    pip install --no-cache-dir . && \
+    pytest && \
+    cd .. && rm -rf ${HOSTILE_VER}.tar.gz
+
+
+# Stage 2: Final image
+FROM ubuntu:focal AS app
+ARG HOSTILE_VER
+
+LABEL base.image="ubuntu:focal" 
+LABEL dockerfile.version="1" 
+LABEL software="hostile"
+LABEL software.version=${HOSTILE_VER}
+LABEL description="Precise host read removal."
+LABEL website="https://github.com/bede/hostile"
+LABEL license.url="https://github.com/bede/hostile?tab=MIT-1-ov-file#readme"
+LABEL maintainer="Taylor K. Paisie"
+LABEL maintainer.email="[email protected]"
+
+COPY --from=builder /usr/ /usr/
+COPY --from=builder /hostile-${HOSTILE_VER}/tests/data/sars-cov-2/ /data/test/sars-cov-2/
+
+RUN apt-get update && apt-get install -y --no-install-recommends \
+    ca-certificates && update-ca-certificates && \
+    apt-get clean && rm -rf /var/lib/apt/lists/*
+
+CMD hostile --help
+WORKDIR /data
+
+# Optional stage: Test data
+FROM app AS test
+
+WORKDIR /data/test
+
+RUN wget https://github.com/bacterial-genomics/test-datasets/raw/assembly/test_data/test_miniburk_R1.fastq.gz \
+    --no-check-certificate && \
+    wget https://github.com/bacterial-genomics/test-datasets/raw/assembly/test_data/test_miniburk_R2.fastq.gz \
+    --no-check-certificate && \
+    hostile clean --index /data/test/sars-cov-2/sars-cov-2 --fastq1 test_miniburk_R1.fastq.gz --fastq2 test_miniburk_R2.fastq.gz
+
+WORKDIR /data

hostile/1.1.0/README.md

@@ -0,0 +1,40 @@
+# hostile
+
+Main tool: [hostile](https://github.com/bede/hostile)
+
+Code repository: https://github.com/bede/hostile
+
+Basic information on how to use this tool:
+- executable: |
+```
+usage: hostile [-h] [--version] {clean,mask,fetch} ...
+
+positional arguments:
+  {clean,mask,fetch}
+    clean             Remove reads aligning to an index from fastq[.gz] input files
+    mask              Mask reference genome against target genome(s)
+    fetch             Download and cache indexes from object storage for use with hostile clean
+
+options:
+  -h, --help          show this help message and exit
+  --version           show program's version number and exit
+```
+
+- help: `hostile --help`
+- version: `hostle --version`
+- description: | 
+> Hostile accurately removes host sequences from short and long read (meta)genomes, consuming single-read or paired fastq input.
+
+
+Full documentation: https://github.com/bede/hostile
+
+
+# Testing hostile analysis
+```
+# Downloading test dataset
+wget https://github.com/bacterial-genomics/test-datasets/raw/assembly/test_data/test_miniburk_R1.fastq.gz --no-check-certificate
+wget https://github.com/bacterial-genomics/test-datasets/raw/assembly/test_data/test_miniburk_R2.fastq.gz --no-check-certificate
+
+# Running hostile clean on test dataset
+hostile clean --index /data/test/sars-cov-2/sars-cov-2 --fastq1 test_miniburk_R1.fastq.gz --fastq2 test_miniburk_R2.fastq.gz
+```