Bestie

WDL based hts-analysis for slurm cluster with enviroment modules. More scalable then ever before setting discrete cpu, memory, runtimes and disk space based on input files for maximum* sheduling efficiency.

*: Always WIP due to lfs stability and edge cases.

Goals:

• is to get generic alignment working.
• basic variant calling (haplotypecallerGvcf).
• somatic variant calling(MuTect2,...).
• ichorCNA integration.
• Variant annotation
• Functional filtering of vcfs
• Stability over > 10 samples
• Rework dir structure to be more in line with warp/other public resources
• End to end easybuild install

Relevant example resources for writing wdl files :

WARP BioWDL

How to use

Run the cromwell validation tool womtool to validate the input and generate a template input file. for sampleJson see ./tests/data/raw/fastq/samples.json. See below for the most simple run example

 java -Xmx8g -Dconfig.file=./path/to/cromwell.conf -jar ./path/to/cromwell.jar run Bestie.wdl -i inputs_integration.json

prepping a samplesheet/json for input

This first part makes a sample tsv with one readgroup per line that can be edited as needed

ls /path/to/raw/fastq/*_R1.fastq.gz | perl -wne 'BEGIN{print "fq1,fq2,sampleName\n"};chomp;print $_;s/_R1./_R2./g; print ",$_"; s/.*/([\w\d]*)_.*/$1/g; print ",$_" ;print "\n"'| perl SampleSheetTool.pl reformatmin /dev/stdin > samplesheet.csv

Converts the samplesheet to json file merging readgroups to samples as needed (some text alignment issues)

perl SamplesheetTool jsondump samplesheet.csv > samplesheet.json

The next command sets up the environment in the /path/to/workflow/output_folder/ copying all the needed files and linking all the needed folders raw and runs cromwell.jar to execute the workflow.

• This needs '-d /path/to/folder/containing/data/' usually the 'apps/' folder for me.
• This results in outputs below '/path/to/workflow/output_folder/' containing all the (fixupped) files needed for running the analysis.

fastq folders (example based on `tests/integration/run_local.sh`):
(
    set -ex
    bash tests/run_project.sh \
        -i $PWD/tests/integration/json/fastqToVariants/inputs_local.json \
        -s samplesheet.json \
        -w $PWD \
        -r /path/to/workflow/output_folder/ \
        -f /path/to/raw/fastq/ \
        -d /path/to/folder/containing/data/
)

How to install

easybuild the required modules or use future wrapper module

Used tools and databases

Name	project website	Article
GNU Parallel	gnu.org	doi
Fastqc	bioinformatics.babraham.ac.uk
BWA	github	preprint
Picard-tools	sourceforge	instructions below faq
GATK4 + MuTect2 toolkit	project home	instructions here
SAMtools	project home	pubmed
HTSeq	project home	pubmed
Cutadapt	github	doi
bcftools	github	pubmed
fgbio	github
freebayes	github	arxiv.org
IchorCNA	github	doi
LoFreq	github	pubmed
MarkTrimming	github
MultiQC	seqera	doi
VEP	github	doi
TrimGalore	github
GATK Bundle	human reference
Ensembl	reference/gtf dowload
UCSC Tools	format conversion/additional tools