search

Illumina / Pisces

Somatic and germline variant caller for amplicon data. Recommended caller for tumor-only workflows.
GNU General Public License v3.0
96 stars 16 forks source link

no allele 2 AD #24

Closed ray1919 closed 5 years ago

ray1919 commented 6 years ago

I found some multi allele position return strange result：

I call with --crushvcf true --ploidy somatic

chr11 33700664 . A C,G 100 PASS DP=530703 GT:GQ:AD:DP:VF:NL:SB 0/1:100:402064,127503:530703:0.2403:30:-100.0000

Why the result have only two AD values while 3 alleles called?

tamsen commented 6 years ago

Hi there,

I don't think we have any use cases for crushvcf on and somatic mode. IMHO its more readable to have one line per allele (uncrushed) especially for somatic and often for germline. "Crushing" into standard vcf format (one line per loci) is only something I (grudgingly) recommend only if you are doing diploid calling (limited ploidy) and you really need it. (And even then, I still dont like the convention of one line per loci)

In your particular case, the two AD correspond to the C and G calls. But I should probably disable crush on + somatic mode. In our group, we never use those two options together, and so the format of the results are not well tested.