Output list of canonical proteome false positive peptides

[lydiayliu]

Given a homozygous SNP / Indel, there shouldn't be a peptide detected that matches the canonical proteome at that location (given that all isoforms that can produce the canonical peptide have the variant). It would be interesting to output such a list of peptides. (One step further is to change the canonical to remove these peptides, but I think most people would rather search with the canonical and post-hoc filter if any).

Currently VEP doesn't include genotype information, I'm not sure if it's possible to have it include it since it is not relevant to the annotation. Would probably require reading in the VCF for genotype information.

[zhuchcn]

Very interesting thought! A thought here is using the idea of VAF. For SNV, homozygous basically means VAF is 100%, and then we can add a filtering step when calling peptides from the PVG graph to remove peptides that don't carry this variant. I assume that this will only be limited to SNV but not any other variant types. Can indel be homosygous?