D3b/COV-IRT RNAseq SNV Calling Workflow (beta)

• Sorting steps removed and duplicates marked directly
• Added explicit bam index input for strelka2 for better index naming style compatibility
• Added very basic VEP r100 annotation tool
• This release was used on cavatica to create "Analysis-ready variants"

D3b/COV-IRT RNAseq SNV Calling Workflow (beta)

This workflow more closely follows how Broad runs in their WDL (no splitting of files nor intervals until haplotype call). Splitting (aside from haplotype calls) during bam pre-processing steps causes differences in overlap of calls and is not yet considered reliable without extensive benchmarking. It follows the COV-IRT pipeline except for gVCF generation. Preparation for joint calls is not part of this workflow and is more of a single sample call style.

D3b RNAseq SNV Calling Workflow (beta)

Updated software to GATK 4.7.1.0. Also added -dont-use-soft-clipped-bases param based on GATK WDL.

D3b RNAseq SNV Calling Workflow (beta)

This repo contains RNAseq SNV calling pipelines using:

• Broad GATK4 best practices
  • The overall workflow picks up from post-STAR alignment, starting at Picard mark duplicates.
  • For the most part, tool parameters follow defaults from the GATK Best Practices WDL, written in cwl with added optimatization for use on the Cavatica platform.
• Strelka2
  • This workflow is pretty straight forward, with a PASS filter step added to get PASS calls.
• VarDict Java
  • This workflow is based on the Vardict run style of BC Bio.