[ensembl-dev] Change in consequence and HGVSp between VEP v99 and v100

[Wallace Ko]

Hello Ensembl Team,

During re-evaluation of some old samples, we found that VEP v99 and v100
give the same HGVSc, but different annotation, including consequence and
HGVSp, e.g.:

   1. HGVSc: NM_001034954.1:c.182T>C
      - v99:
         - HGVSp: NP_001030126.1:p.Pro61=
         - Consequence: synonymous_variant
      - v100:
         - HGVSp: NP_001030126.1:p.Leu61Pro
         - Consequence: missense_variant
      2. HGVSc: NM_000283.3:c.928-9_940dup
   - v99 HGVSp:  NP_000274.2:p.Tyr314CysfsTer50
      - v100 HGVSp:  -
   3. HGVSc: NM_001040108.1:c.4243-36_4243-1dup
      - v99 HGVSp:
       NP_001035197.1:p.Gln1414_Ile1415insHisIleAlaAsnTerSerTyrSerPhePheProTer
      - v100 HGVSp:  NP_001035197.1:p.Gln1414_Ile1415insHisIleAlaAsnTer

I tried using offline VEP v100 with v99 cache and vice versa. VEP v100
produces the new results regardless of the cache being used. So it should
be due to code change instead of data update.

The change is sort of significant but there seem no relevant details in the
release descriptions (
https://github.com/Ensembl/ensembl-vep/releases?q=release%2F100) or the
mail list about the change.

I wonder what causes such changes and if there is a scope or pattern on the
changes. Otherwise, we may need to re-analyze all results from VEP v99 or
older.

Thank you.

Regards,
Wallace Ko
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ensembl.org/pipermail/dev_ensembl.org/attachments/20220120/011e8c23/attachment.html>