WDL Canales

Bioinformatic pipelines written in the Workflow Description Language (WDL)

Assemble

Requirements:

• python2 or python3

The workflows are split into multiple files that need to be assembled into WDL files by

make

Now, WDL files can be found in wdl/.

Run the workflows

The workflows can be run using Cromwell (command line) or Firecloud (web service hosted on the Google Cloud platform). See their respective documentations for further details.

The workflows here are also released under the dshih namespace on Firecloud.

Test the workflows

The workflows are continuously tested on pre-generated test data on Travis.

To run these tests yourself, you need to first fulfill the following requirements:

• java (jre8)
• docker

The workflows can be run on pre-generated test data using cromwell (automatically retrieved). Simply do

make check

The exit codes of each task is printed to stdout.

Generate test data

Requirements:

• dgswim
• samtools
• bwa

You can re-generate the test data by

cd test
./make.sh

Workflows

Align reads from pair-end FASTQ:

• fastq_align_paired

Extract reads from pair-end BAM and re-align to another reference:

• bam_realign_paired
• bam_realign_paired_fast
• bam_realign_paired_faster

Filter out OxoG and FFPE artifacts from SNVs:

• maf_filter_orient_bias_legacy