Wrong HGVSc when using RefSeq transcripts

[jpuntomarcos]

Describe the issue

HGVSc annotations are not correct when the NM sequence differs from the GRCH37 sequence. For example, If I use VEP online (GRCH37) to query 11:64572018_A/C variant with refseq transcripts, I get:

HGVSp and codons are correct: In this position, the NM transcript contains a G and the alternate allele provided is a T (-strand), so predicted GCA/TCA codons are right (and the HGVSp also). However, the HGVSc notation seems wrong: NM_000244.3:c.1636A>T is actually the one we can obtain when annotating against Ensembl transcripts or using the --use_given_ref parameter to use the ref allele provided by the user.

Thanks :)

Additional information

System

• VEP version: 104 online
• Link to online job

Full VEP command line

./vep --appris --buffer_size 500 --check_existing --distance 5000 --hgvs --mane --polyphen b --refseq --regulatory --sift b --species homo_sapiens --symbol --transcript_version --tsl --cache --input_file [input_data] --output_file [output_file] --port 3337

[dglemos]

Hi @jpuntomarcos,
In cases where RefSeq transcripts do not match the reference genome, we are using alignment files provided by NCBI to create a new reference, matching the transcript and use this for consequence calling. The HGVS calculation does not currently use this reference modification.

[jpuntomarcos]

Thanks @dglemos . Is it planned to modify this to use the NCBI reference?

[dglemos]

We have a plan to improve the HGVS calculation, however I can't give you an exact date when it will be done.

[bibinf]

Hi @dglemos ,

I'm not quite sure if my issue is related to this one, but I've also some strange behavior with a bam-edited annotation:

http://grch37.ensembl.org/Homo_sapiens/Tools/VEP/Results?tl=yJr1DXAflVyUlOCG-7819531

The calculated HGVS.c and HGVS.p states synonymous_variant, NM_000130.5:c.1601A>G, NP_000121.2:p.Arg534=, but I think it should be missense_variant, c.1601G>A, p.Arg534Gln after the correction took place or am I wrong?

You wrote that the HGVS calculation is not using this reference modification, but on the other hand you see an Arg as reference amino acid. Otherwise it would have been Gln.

More to the background:
The variant is heterozygous in our case. Hence, after your correction the "p.Arg534=" only fits to the "new wildtype" allele and not the allele with the actual variant.

Best regards and thanks in advance,
Sebastian

[dglemos]

Hi @bibinf,
The HGVSc is not using the reference modification as it should, see your example NM_000130.5:c.1601A>G.
You are correct, in your example the amino acid change should be p.Arg534Gln (G>A and not G>G). This issue needs to be addressed - thank you for reporting it.

Best wishes,
Diana