chunjie-sam-liu/rmats-turbo-wdl

rmats-turbo-wdl

rMATS turbo v4.1.2 WDL workflow for AnVIL

This is the rMATS turbo workflow in Workflow Description Language (WDL) written to run on the AnVIL cloud platform. Bam files are taken as input.

Dependencies

This workflow is written to run on the AnVIL platform powered by Terra. In order to run locally, cromwell and womtool must be installed. Run the workflow locally using:

java -jar cromwell-xx.jar run rmatsTurbo.wdl -i xx.json

Example

Generate input file (xx.json)

Prepare the test data, and write them into the xx.json file. The file can be generated with suggested formatting using womtool, e.g.,

java -jar womtool-xx.jar inputs rmatsTurbo.wdl > test.json

The output .json file shows all the defined parameters (required and optional), the data type and default value if optional.

{
  "rMATS_turbo.mil": "Int (optional, default = 50)",
  "rMATS_turbo.variable_readLength": "Boolean (optional, default = false)",
  "rMATS_turbo.paired_stats": "Boolean (optional, default = false)",
  "rMATS_turbo.novelSS": "Boolean (optional, default = false)",
  "rMATS_turbo.nthread": "Int (optional, default = 1)",
  "rMATS_turbo.statoff": "Boolean (optional, default = false)",
  "rMATS_turbo.readLength": "Int",
  "rMATS_turbo.lib_type": "String (optional, default = \"fr-unstranded\")",
  "rMATS_turbo.pairedORsingle": "String (optional, default = \"paired\")",
  "rMATS_turbo.tmp_dir": "String",
  "rMATS_turbo.gtf": "File",
  "rMATS_turbo.anchorLength": "Int? (optional)",
  "rMATS_turbo.bam_g2": "Array[File]",
  "rMATS_turbo.tstat": "Int (optional, default = 1)",
  "rMATS_turbo.mel": "Int (optional, default = 500)",
  "rMATS_turbo.task_type": "String (optional, default = \"both\")",
  "rMATS_turbo.bam_g1": "Array[File]",
  "rMATS_turbo.out_dir": "String",
  "rMATS_turbo.cstat": "Float (optional, default = 0.0001)"
}

Then open the .json file in editor, keep the keys unchanged, and fill in the values each key is expecting. (optional) parameters can be omitted and default values will be used. Parameter documentation can be checked in the rmatsTurbo.wdl under the parameter_meta sections.

{
  "rMATS_turbo.bam_g1": ["../testdata/231ESRP.25K.rep-1.bam", "../testdata/231ESRP.25K.rep-2.bam"],
  "rMATS_turbo.bam_g2": ["../testdata/231EV.25K.rep-1.bam", "../testdata/231EV.25K.rep-2.bam"],
  "rMATS_turbo.gtf": "../testdata/gtf/Homo_sapiens.Ensembl.GRCh37.75.gtf",
  "rMATS_turbo.pairedORsingle": "paired",
  "rMATS_turbo.readLength": 50,
  "rMATS_turbo.nthread": 4,
  "rMATS_turbo.tmp_dir": "rmats_tmp",
  "rMATS_turbo.out_dir": "rmats_out"
}

Run the wdl workflow

Run the wdl workflow locally:

java -jar cromwell-xx.jar run rmatTurbo.wdl -i test_bam_default.json

OR submit the xx.json file to AnVIL under the [INPUTS] tab of rmats_turbo_bam workflow. Users can write the Attribute column using json format or click on upload json on the upper-right corner of INPUTS section. NOTE that the /path/to/test_data should be modified to be data path on AnVIL cloud.

Save the editting, and click on RUN ANALYSIS. The job will be submitted with a link of job manager to check for task status. Using the above .json as input, the outputs are compressed into a .tar file in the output directory:

rmats_out.tar