A CCB researcher emailed about how 17:7574012 G>A has HGVS of ENST00000269305.4:c.1015G>T
As it's -'ve strand you'd think given the ref G would be complement C ie HGVS should be ENST00000269305.4:c.1015C>T
It looks like VEP calculates HGVS using the fasta reference base, this kind of makes sense (due to normalization and annotations etc) but I have raised an issue with them to make it more explicit in their documentation: https://github.com/Ensembl/ensembl-vep/issues/865
In the mean time, I think we should definitely add a warning about the reference base being different than our Fasta and what that means. We could initially use PyHGVS and long term modify our annotation to add a CHECK_REF option which we can display a warning about on the variant details page:
A CCB researcher emailed about how 17:7574012 G>A has HGVS of ENST00000269305.4:c.1015G>T
As it's -'ve strand you'd think given the ref G would be complement C ie HGVS should be ENST00000269305.4:c.1015C>T
It looks like VEP calculates HGVS using the fasta reference base, this kind of makes sense (due to normalization and annotations etc) but I have raised an issue with them to make it more explicit in their documentation: https://github.com/Ensembl/ensembl-vep/issues/865
In the mean time, I think we should definitely add a warning about the reference base being different than our Fasta and what that means. We could initially use PyHGVS and long term modify our annotation to add a CHECK_REF option which we can display a warning about on the variant details page:
https://asia.ensembl.org/info/docs/tools/vep/script/vep_options.html#dont_skip
Combining --check_ref with --dont_skip will add a CHECK_REF output field when the given reference does not match the underlying reference sequence.