TODO: v1.0.0 Release

[skchronicles]

Germline Pipeline

• Only run DV and its subsequent rules with norm samples
  • If pairs file (only relevant to the somatic pipeline, i.e. provided --call-somatic), then pull out normal samples
  • If --call-somatic is not provided then all samples are assumed to be normals
  • If --call-somatic is provided and no pairs file, then all samples are assumed to be tumors

This will prevent any unnecessary compute and will reduce overall runtime.

Fix strelka issue

Filtering from GATK is not working for strelka, resolved by adding these two commands before it goes to norm & splitting:

bcftools concat -Ov -a  \
    -D somatic.snvs.vcf.gz somatic.indels.vcf.gz \
    -o strelka.merge.indel.snps.vcf
java -Xmx16g -cp /data/OpenOmics/references/genome-seek/hmftools/purple_v3.2.jar com.hartwig.hmftools.purple.tools.AnnotateStrelkaWithAllelicDepth -in strelka.merge.indel.snps.vcf -out strelka.merge.indel.snps.annotated.vcf.gz

Next fix is for spliting TUMOR from strelka. When splitting we cannot use "-c1" for strelka as we don't have the tag in vcf file to check for --min-ac/--max-ac. We need to update the tumor splitting command for strelka to remove -c1:

bcftools view -s TUMOR  -Oz -o strelka.tumor.vcf.gz \
    strelka.merge.indel.snps.annotated.filtered.vcf.gz

Tumor only

Need to update rule all so that the strelka and muse output files do not get created for tumor-only samples