Exome variant filtering questions

[Bio_Stacker_300]

Hey, I'm checking out exome data from folks with a rare disease and trying to improve how I filter stuff. I know about read depth and Phred scores, but I'm a bit stuck on things like strand bias, how GC content and homopolymer runs near variants affect stuff, mapping quality, and variant confidence. Any tips on when to use these filters or good thresholds? I'd love to hear how you handle this in your pipelines

 

[TropicalMe]

When I'm filtering exome data for rare variants, I usually treat those metrics as layers of confidence rather than hard cutoffs. Strand bias can flag variants that are only seen on one read direction, if it's extreme, it's usually a red flag. GC content and homopolymers can cause sequencing errors, so variants in long homopolymer runs or extreme GC regions deserve extra scrutiny.