search

HKU-BAL / ClairS

ClairS - a deep-learning method for long-read somatic small variant calling
BSD 3-Clause "New" or "Revised" License
67 stars 7 forks source link

Germline variants present in output.vcf #12

Closed xingyaoc closed 1 year ago

xingyaoc commented 1 year ago

According to the ClairS documentation in the README, I expect the germline variants to be eliminated in the somatic vcf, and this is consistent with the demo outputs. However, running ClairS on my own data, I see overlaps between the clair3 germline callsets (both normal and tumor) and the final somatic callset. Is this behavior expected? If so, can you provide guidance on how to handle the overlapping calls?

aquaskyline commented 1 year ago

Clair3 calls variants with AF down to 0.15. It is being called a germline variant caller but that doesn't mean variants called by Clair3 have to be germline.