Mutation detection in RNA-Seq highlights the GATK Best Practices in RNA-Seq variant calling, several sources of variant annotation, filtering based on CRAVAT, and interactive data exploration using IGV web. Initially a 2-pass STAR alignment is performed followed by a series of GATK tools for variant calling. To aid in interpretation, variants are further annotated using several RNA editing databases (DARNED, RADAR), cancer specific features (COSMIC), and general variant information (dbSNP, SnpEff). CRAVAT is then used to score variants as cancer drivers (CHASM scores) or pathogenic (VEST scores). The mutation detection system adds value to GATK variant calling with tiered filtering. Initially, a comprehensive VCF file with conservative filtering is created. A VCF file additionally filtered based on annotation sources is then provided (for example filtering known RNA editing events and common variants). Final filtering is focused on reducing variants to a short list of predicted cancer-focused variants (which is made available as a VCF file). To aid users in data exploration, an interactive IGV web browser has been integrated directly into Galaxy giving an immediate customized view of the final cancer-focused variants.
For more information: https://github.com/NCIP/Trinity_CTAT/wiki