Open chulghim opened 2 years ago
Generally the genetic map does't make a large difference, as long as the reference panel is large enough (so switching among reference haplotypes while imputing doesn't occur very often). In case of a poor fit between reference panel populations and target population to be imputed, I would recommend a diverse population if possible, or a map derived from an ideally large population which substantial African ancestry in which many PRDM9 alleles are segregating (e.g. ASW).
I don't have data on HRC vs 1kgp3 panel. The latter is included in the former, so I suspect the HRC will almost certainly do better, though I don't know the precise difference. The HRC paper probably contains results that generally showcase how much more accurate it is than 1kgp3.
1.In case of mixed race, how can make genetic_map file?
2. It seems that HRC panel was used in this paper, I wonder that HRC panel vs 1kpg3 panel's imputation accuracy difference.