Dear all,
I have a question regarding the VEP --per_gene
option.
In the manual is explained that using this option the "worst" mutation is chosen for every variant position on every gene.
Hence if the base in the reference is C and there are two transcripts TrX and TrY and two alternative allele A and T, with this option VEP chose first the best transcript (e.g. Tx) and then the allele that generate the worst outcome on that transcript (e.g. A).
I wonder, in case of "similar" outcome by the two different alleles (e.g. they cause both a missense mutation), if the Allele Frequency is taken in consideration for the choice.
Thanks a lot.
I always recommend decomposing variants into biallelic entries for precisely this reason. I think VEP will pick arbitrarily when faced with the scenario you describe.