Can we replace LDSC with something like HDL to give more accurate results for smaller datasets or is GSEM from smaller cohorts (or Z<4) unreliable too?
Replying to “Can we replace LDSC with something like HDL to giv…”:
HDL works better with smaller sample sizes indeed
Replying to “Can we replace LDSC with something like HDL to giv…”:
Yes you can - there genomic SEM package has been extended to use HDL estimates instead of LDSC - see this tutorial RPubs - HDL in GenomicSEM