Galaxy - RNA Seq workflow misc. issues

[CaseyRichards92]

Add issues or preferences for RNA Seq workflow https://hardwoodgenomics.org/content/whole-thing-617

[CaseyRichards92]

Step 2

• Change sentence "Only files of the following types are listed: fastqsanger, fastqsanger.gz, fastqillumina, fastqillumina.gz, fastqsolexa, fastqsolexa.gz." to "Only .fastq and .fastq.gz files accepted from the following platforms illumina, sanger, solexa."
• Change to Paired end files opposed to Paired files
• Doesn’t seem to accept fastq or fastq.gz files
• Reword description under Data Collection: “You may select from previously uploaded files, or upload new files.”
• Todo: 64 GB is a lot of data to allow for each user.

[mestato]

Alteration from above:

• Change sentence "Only files of the following types are listed: fastqsanger, fastqsanger.gz, fastqillumina, fastqillumina.gz, fastqsolexa, fastqsolexa.gz." to "Only .fastq and .fastq.gz files accepted from the Illumina platform"

[CaseyRichards92]

Galaxy workflow link https://treegenes.cam.uchc.edu/workflows/run?id=96d9e11f37f34b29

[CaseyRichards92]

Featured counts from test

[mestato]

• Stringtie step - Add a note to users that they can use gtf and just rename it gff (yes this is a hack but is fast and will work)

[florence-77]

User data quota lowered to 20GB. Is this large enough to be useable, but small enough to prevent the site from being overloaded?

[mestato]

Update tool descriptions:

• StringTie merge (version 2.1.1): Merge transcripts across samples into a nonredundant set
• featureCounts (version 1.4.6.p5): Count reads per transcript

Also:

• grey out presets using css
• indent lines under the bolded headings (also a CSS thing)

[Ferrisx4]

Workflow Logic

Option 1 (Preferable, but not guaranteed to work)

• Allow Tripal Galaxy to read from Galaxy when tools specify that they want a Genome from the site (Data managers)
  • This will allow us to take advantage of the Galaxy built-in Genomes and their respective index files. All workflows in the future should be this way. We already have indexes for many types, all we need to do is add them to the .loc files in Galaxy. See the diagram on this page for an explanation.

OR

Option 2 (Should work, more time needed to re-wrap hisat and to wrap hisat-build)

• Allow the HISAT2 tool to accept indexed reference genomes from history
• Wrap hisat-build as a tool, insert it into Workflow before HISAT
  • This is to avoid having HISAT re-index the reference genome for every pair of input files (bad)

[Ferrisx4]

Modified Galaxy webform logic to accept certain file aliases but also a compressed version of any file: tripal/tripal_galaxy@d77d14b

This is necessary because Drupal seems to be unable to filter correctly on multipart file extensions.
Ideally we can trust our users to be able to read the form and supply the correct type of file.