Skip to content

Latest commit

 

History

History
32 lines (25 loc) · 1.26 KB

README.md

File metadata and controls

32 lines (25 loc) · 1.26 KB
Raw

Automated Pipeline based on Snakemake

This workflow expects SomaticSeq and the desired suite of variant callers to be installed. That is, their executables should be found via the PATH environment variable.

For SomaticSeq, it is sufficient to include the top-level directory and the utilities directory:

SOMATICSEQ_HOME=/path/to/somaticseq
export PATH=$SOMATICSEQ_HOME:$SOMATICSEQ_HOME/utilities:$PATH

Example usage:

snakemake \
    -j \
    --config \
        tumor=/ABSOLUTE/PATH/TO/tumor.bam \
        normal=/ABSOLUTE/PATH/TO/normal.bam \
        reference=/ABSOLUTE/PATH/TO/GRCh38.fa \
        dbsnp=/ABSOLUTE/PATH/TO/dbSNP.GRCh38.vcf \
        gatk=/ABSOLUTE/PATH/TO/GATK.jar \
        varscan=/ABSOLUTE/PATH/TO/VarScan.jar \
        caller_threads=36 \
    somaticseq

caller_threads is the number of threads to be used for each of the variant callers that support parallelization.

The config.yaml file specifies default options, mostly for specifying which variant callers' results you'd like to feed into SomaticSeq. You may pass those options on the command line, as is done for caller_threads above, and whatever is passed on the command line will override what is specified in the configuration file.