Compare mehari annotation with VEP annotation

[tedil]

• build mehari transcript DB (ensembl):
  • download cdot for ensembl and grch37 / grch38
  • download ensembl FASTA for transcripts
  • create / fill seqrepo with ENSEMBL FASTAs
  • build mehari database
• install VEP + caches
  • caches: grch37 release 105, grch38 release 110
• obtain test dataset, e.g.:
  • clinvar VCF
    • https://ftp.ncbi.nlm.nih.gov/pub/clinvar/vcf_GRCh37/
    • https://ftp.ncbi.nlm.nih.gov/pub/clinvar/vcf_GRCh38/
  • genes with few variants: TGDS, KYNU
  • genes with lots variants: BRCA1, BRCA2, TTN
  • later: extend to include certain regions in gnomAD (via tabix https://)
    • regions: BRCA1, BRCA2, TTN, SLC39A14 (ManePlusClinical)
• comparison:
  • VEP on ClinVar GRCh37/GRCh38
  • local check: build data for e.g. TGDS
  • annotate all transcripts with both mehari and vep

[xiamaz]

I had started https://github.com/varfish-org/annotation-zoo a while ago, but that repo got stalled, since I was busy with DHA stuff. It would be great if we could put repro stuff there unless we want to move towards a more monorepo approach for mehari.

[holtgrewe]

I had started https://github.com/varfish-org/annotation-zoo a while ago, but that repo got stalled, since I was busy with DHA stuff. It would be great if we could put repro stuff there unless we want to move towards a more monorepo approach for mehari.

Let's continue in a central repo, but I'd rather have this called mehari-validation or similar?

[tedil]

We do this in https://github.com/varfish-org/mehari-annotation-comparison but this potentially needs some tidying up and a README