workflow.cwl

#!/usr/bin/env cwl-runner

doc: |
  DESeq2 example workflow for **differential gene expression analysis**
  
  This workflow runs DESeq2 on the output of the kallisto workflow
  and the metadata file.
  It runs an R script, deseq2.R, which ideally should be split into three sub scripts and accordingly three workflow steps
    1. Read kallsito data
    2. Prep / run deseq2
    3. Plot results

  ## DESeq2 docs:
    https://bioconductor.org/packages/release/bioc/html/DESeq2.html

  ## Importing kallisto output with tximport
    https://bioconductor.org/packages/release/bioc/vignettes/tximport/inst/doc/tximport.html#kallisto

  ## Multi-package containers
  - R and combinations of library dependencies are available as multi-package containers from [BioContainers](https://github.com/BioContainers/multi-package-containers)
  - Searched for `repo:BioContainers/multi-package-containers deseq2 tximport rhdf5`
  - and found `quay.io/biocontainers/mulled-v2-05fd88b9ac812a9149da2f2d881d62f01cc49835:a10f0e3a7a70fc45494f8781d33901086d2214d0-0` :tada:

cwlVersion: v1.2
class: CommandLineTool
hints:
  DockerRequirement:
    dockerPull: quay.io/biocontainers/mulled-v2-05fd88b9ac812a9149da2f2d881d62f01cc49835:a10f0e3a7a70fc45494f8781d33901086d2214d0-0
requirements:
  - class: InitialWorkDirRequirement
    listing:
      - entryname: deseq2.R
        entry:
          $include: deseq2.R
  - class: NetworkAccess
    networkAccess: true
baseCommand: [Rscript, deseq2.R]
inputs:
  inKallistoResults:
    type: Directory
    inputBinding:
      position: 1
  inMetadataFile:
    type: File
    inputBinding:
      position: 2
  inMetadataSample:
    type: string
    inputBinding:
      position: 3
  inMetadataFactorList:
    type: string[]
    inputBinding:
      position: 4

outputs:
  output:
    type: File[]
    outputBinding:
      glob:
        - "*.svg"
        - "*.csv"