ComparativeGenomicsToolkit / cactus

Official home of genome aligner based upon notion of Cactus graphs
Other
526 stars 111 forks source link

Turn vcf left-shifting off by default. #1479

Closed glennhickey closed 2 months ago

glennhickey commented 2 months ago

I added bcftools norm -f (left shifting) to the VCF export logic by default in 2.8.2. This makes for more consistent variant representation, but can result in overlapping sites. This apparently upsets pangenie, for which I don't want to break support.

So this PR backtracks on this, making bcftools norm off by default. Someone wanting to run it can do so pretty easily themselves, or turn it back on in the configuration. It's still on by default for output normalized by vcfwave.

There's a vcflib tool called vcfcreatemulti that appears to merge overlapping variants. I'm a bit reluctant to try running it by default now because I'm not sure what it does to tags and, crucially, I don't have a way right now of putting it into the binary release. People using the Cactus docker image can try it out if they want as it's pretty easy to run.

Resolve #1477 and #1460