BRCA Variant with Conflicting CAVA and other annotation results

[Sally Guthrie]

I ran CAVA on the 1000 Genomes whole genome sequences (ftp://ftp-trace.ncbi.nih.gov/1000genomes/ftp), and found that HG02147 had a complex variant that CAVA and other annotation pipelines annotated differently. I thought this might be an interesting variant for CAVA developers.

The variant:
#CHROM POS  ID REF ALT QUAL FILTER INFO 

13 32913957 . ATA TAT .        . AN=2;AC=1;CGA_FI=675|NM_000059.3|BRCA2|CDS|NONSENSE;TYPE=Complex;ENST=ENST00000380152;GENE=BRCA2;TRINFO=+/83.7kb/27/10.9kb;LOC=Ex11;CSN=c.5465_5467delinsTAT_p.Asn1822_Tyr3417del;CLASS=IF;SO=.;IMPACT=2;ALTANN=.;ALTCLASS=.;ALTSO=.;DBSNP=.

It appears CAVA annotated this variant as an inframe insertion/deletion.

Annotating this genome with GET-Evidence, an annotation pipeline developed by the Harvard Personal Genome Project, this variant is classified as a missense followed by a stop-gain mutation (NK1822I*).

[Elise Ruark]

Thank you very much for reporting this. It does appear to be a bug for indels which result in a stop and also change the amino acid sequence upstream of the new stop codon. We are now working on fixing this so that the protein level annotation is consistent with the nomenclature, i.e. p.Asn1822_Lys1823delinsIleX. The class will also then become SG (stop-gain) with the correct SO term.