FFPE and cfDNA evaluations

[davidbenjamin]

Since most of the work is in setting up the necessary tools and pipelines to evaluate, I will lump the actual act of evaluating on specific data into this single ticket. We need to:

• CRSP specificity: apply Takuto's mutect2-replicate-validation.wdl on the CRSP NA12878 replicates.
• CRSP sensitivity: apply the (currently in-progress) hapmap sensitivity pipeline to CRSP data.
• cfDNA: run cfDNA samples and matched solid tumor samples (which we already have from Viktor) and run the concordance tool.
• FFPE: run FFPE and matched non-FFPE samples and run the concordance tool.
• tumor-only: run some TCGA samples with and without their matched normal and run the concordance tool.

[davidbenjamin]

Update: CRSP sensitivity and specificity have been run several times, cfDNA is currently running for the first time. FFPE and tumor-only will use the same wdl as cfDNA, so we'll run those once cfDNA finishes.

[davidbenjamin]

Update: everything done except FFPE and tumor-only. FFPE will use the same wdl as cfDNA.

[droazen]

Issue moved to broadinstitute/gatk #2942 via ZenHub