Closed ptrebert closed 2 years ago
this has been implemented and tested for the first handful of samples. so far, looks like we get <100 HET SNPs per assembly (irrespective of population of origin) for HiFi with Qual >= 10 For ONT/HG00358, no variant left after filtering, more test samples still running...
@ptrebert Sorry, forgot to reply to this. Sounds pretty good! I bet there is clustering of these SNPs to specific regions. Based on HiFi depth there are some collapses in some samples, at least some are probably located in those regions as well.
identify clusters of HET variants as potential misassembled regions