Replace MetaLR and Mutation Assessor with AlphaMissense and ESM1b - 116

[nakib103]

Description

We show several pathogenicity scores from different tools in variation pages. In this PR we have removed MetaLR and Mutation Assessor as they are bit old and added AlphaMissense and ESM1b scores which are mote state-of-the-art.

Views affected

1. variation page - http://wp-np2-35.ebi.ac.uk:7070/Homo_sapiens/Variation/Mappings?db=core;g=ENSG00000175324;r=8:38163339-38176493;t=ENST00000311351;v=rs148018686;vdb=variation;vf=545300927#ENST00000520755_545300927_G_tablePanel
2. variant table in transcript - http://wp-np2-35.ebi.ac.uk:7070/Homo_sapiens/Transcript/Variation_Transcript/Table?db=core;g=ENSG00000175324;p=ENSP00000310596;r=8:38163339-38176493;t=ENST00000311351
3. variants in protein - http://wp-np2-35.ebi.ac.uk:7070/Homo_sapiens/Transcript/ProtVariations?db=core;g=ENSG00000175324;r=8:38163339-38176493;t=ENST00000311351

Possible complications

Not that I am aware of.

Merge conflicts

No conflicts

Related JIRA Issues (EBI developers only)

https://embl.atlassian.net/browse/ENSVAR-6812

[jamie-m-a]

This generally looks great - thank you!

A couple of things to double check before I approve:

• Rounding to 1dp not always working for some esm-1b scores, try truncating if not rounded properly? Or another maths module maybe?
• About half the missense variants in the variant table have no revel/alphamissense/esm1-b but do have CADD. make sure this is expected from data used for sandbox (as scores are present in dbNSFP file)

[jamie-m-a]

One more thing - the protein variations view seems to be missing the ESM1-b scores:
http://wp-np2-35.ebi.ac.uk:7070/Homo_sapiens/Transcript/ProtVariations?db=core;g=ENSG00000175324;r=8:38163339-38176493;t=ENST00000311351

[jamie-m-a]

Issues mentioned in comments all resolved - thanks Nakib!