Repository metrics
- Stars
- (0 stars)
- PR merge metrics
- (No merged PRs in 30d)
Description
One suggestion that @rmccoy7541 had brought up in a discussion was that while Uniparental isodisomy is undistinguishable from meiotic monosomy on the basis of B-allele frequency data alone, we might have decent power to detect hetero-disomy (both haplotypes from a single parent copied).
The hidden states would be (0,1,-1,-1) and (-1,-1,0,1) in the MetaHMM class. Since these are fairly easy to add in the model as currently written we can add those in. I think that the karyotype labels would be 2m and 2p for the output of this.
This feature is currently de-prioritized as we don't expect it to add to phenotypes that are valuable for GWAS (e.g. uniparental disomy is typically quite rare, even in embryos). But we can make sure that this is a feature that is potentially toggleable in MetaHMM.