HGVSg and HGVSc notation generated by the REST API response are incorrect for some variants

[Karthik-Tempus]

Usually HGVSg notations from Variation/VEP REST API are consistently shifted to the furthest 3' position in the chromosome's forward orientation. However, for some variants we noticed that the HGVSg notation from Variation REST API response is only partially shifted, not shifted to the furthest 3' position.

For instance, this variant: 17:29508461:C:CCCTAAAGAAGGTTGCGCAGTTAGCAGTTATAAATAGCCTGGAAAAGGTAAGTTACAACCTCTCTGGTATTAAAATTTTGTTTTTGATGTAAAATTTGCTGTTGTTAGCATCCTGAATCAAAAAGTTATGACTTGAGTGATAGTTTCACATTCATTTTCAGGAAGAATACATTGTAATATTATTATGAAGGAAGTTAGAAGTTTGTGACATTTTATTTACTGTATTACAAAAAATCACTGTAAAGACATGTGGTTCTTTATTTATAGGCATTTTGGAACTGGGTAGAAAATTATCCAGATGAATTTACAAAACTGTACCAGATCCCACAGACTGATATGGCTGGTAAT

When this variant is queried using this hyperlinked URL, the HGVSg notation generated is as follows:
"hgvsg":"NC_000017.10:g.29508512_29508513insTTACAACCTCTCTGGTATTAAAATTTTGTTTTTGATGTAAAATTTGCTGTTGTTAGCATCCTGAATCAAAAAGTTATGACTTGAGTGATAGTTTCACATTCATTTTCAGGAAGAATACATTGTAATATTATTATGAAGGAAGTTAGAAGTTTGTGACATTTTATTTACTGTATTACAAAAAATCACTGTAAAGACATGTGGTTCTTTATTTATAGGCATTTTGGAACTGGGTAGAAAATTATCCAGATGAATTTACAAAACTGTACCAGATCCCACAGACTGATATGGCTGGTAATCCTAAAGAAGGTTGCGCAGTTAGCAGTTATAAATAGCCTGGAAAAGGTAAG"

This HGVSg puts the insertion at this position 29508512_29508513. However, the furthest 3' shifted position would be 29508807_29508808. So, the correct HGVSg notation would be as follows:
NC_000017.10:g.29508807_29508808insTCCTAAAGAAGGTTGCGCAGTTAGCAGTTATAAATAGCCTGGAAAAGGTAAGTTACAACCTCTCTGGTATTAAAATTTTGTTTTTGATGTAAAATTTGCTGTTGTTAGCATCCTGAATCAAAAAGTTATGACTTGAGTGATAGTTTCACATTCATTTTCAGGAAGAATACATTGTAATATTATTATGAAGGAAGTTAGAAGTTTGTGACATTTTATTTACTGTATTACAAAAAATCACTGTAAAGACATGTGGTTCTTTATTTATAGGCATTTTGGAACTGGGTAGAAAATTATCCAGATGAATTTACAAAACTGTACCAGATCCCACAGACTGATATGGCTGGTAA

Then if we take the partially 3' shifted HGVSg notation generated by the Variation REST API and query the VEP REST API endpoint using this hyperlinked URL, the HGVSc notations generated are at the furthest 3' shifted position however the inserted sequence in the HGVSc is a mix, one of the transcripts got the partially shifted sequence and the other two got the completely shifted sequence, see the inserted sequence differences between the transcripts:

"hgvsc": "NM_000267.3:c.730+4_730+5insTTACAACCTCTCTGGTATTAAAATTTTGTTTTTGATGTAAAATTTGCTGTTGTTAGCATCCTGAATCAAAAAGTTATGACTTGAGTGATAGTTTCACATTCATTTTCAGGAAGAATACATTGTAATATTATTATGAAGGAAGTTAGAAGTTTGTGACATTTTATTTACTGTATTACAAAAAATCACTGTAAAGACATGTGGTTCTTTATTTATAGGCATTTTGGAACTGGGTAGAAAATTATCCAGATGAATTTACAAAACTGTACCAGATCCCACAGACTGATATGGCTGGTAATCCTAAAGAAGGTTGCGCAGTTAGCAGTTATAAATAGCCTGGAAAAGGTAAG"

"hgvsc": "NM_001042492.3:c.730+4_730+5insTCCTAAAGAAGGTTGCGCAGTTAGCAGTTATAAATAGCCTGGAAAAGGTAAGTTACAACCTCTCTGGTATTAAAATTTTGTTTTTGATGTAAAATTTGCTGTTGTTAGCATCCTGAATCAAAAAGTTATGACTTGAGTGATAGTTTCACATTCATTTTCAGGAAGAATACATTGTAATATTATTATGAAGGAAGTTAGAAGTTTGTGACATTTTATTTACTGTATTACAAAAAATCACTGTAAAGACATGTGGTTCTTTATTTATAGGCATTTTGGAACTGGGTAGAAAATTATCCAGATGAATTTACAAAACTGTACCAGATCCCACAGACTGATATGGCTGGTAA"

"hgvsc": "NM_001128147.3:c.730+4_730+5insTCCTAAAGAAGGTTGCGCAGTTAGCAGTTATAAATAGCCTGGAAAAGGTAAGTTACAACCTCTCTGGTATTAAAATTTTGTTTTTGATGTAAAATTTGCTGTTGTTAGCATCCTGAATCAAAAAGTTATGACTTGAGTGATAGTTTCACATTCATTTTCAGGAAGAATACATTGTAATATTATTATGAAGGAAGTTAGAAGTTTGTGACATTTTATTTACTGTATTACAAAAAATCACTGTAAAGACATGTGGTTCTTTATTTATAGGCATTTTGGAACTGGGTAGAAAATTATCCAGATGAATTTACAAAACTGTACCAGATCCCACAGACTGATATGGCTGGTAA"

Please consider fixing these bugs as soon as possible.

[dglemos]

Hi @Karthik-Tempus,
Thanks for the query. We'll look into it, I'll let you know when there are any updates.

Best wishes,
Diana