search

brentp / duphold

don't get DUP'ed or DEL'ed by your putative SVs.
MIT License
100 stars 9 forks source link

SNP annotation filters #16

Closed dmckean closed 5 years ago

dmckean commented 5 years ago

Annotating with --snp FILE might add a step in getting a potentially large BCF w/genotypes ready for duphold. Does it make sense to add another option for filtering the SNP BCF that will be used for DHET and DHHU annotations?

Adding a full bcftools-style --include option w/ argument may be overkill. A step down could be to include --filter FLT0[;FLT1...] or just --pass.

I believe a duphold removes indels from a SNP BCF, drops genotypes from other samples to speed up parsing, and drops SNPs where QUAL < 20. If adding an option doesn't make sense, adding v.FILTER == "PASS" might be a good default for most uses.

brentp commented 5 years ago

I pushed a change to only consider PASS snps. It will be out in the next release. Thanks for pointing this out.

brentp commented 5 years ago

... though I should add, I don't really recommend using --snp unless you have a specific reason. Change in depth is a better metric.