Bjarni J. Vilhjalmsson

Results 3 comments of Bjarni J. Vilhjalmsson

Running LDpred2 on COJO residualized summary stats is maybe not a very common idea, but a good idea in this situation. Using COJO you might have to account for a...

Yes, true. It becomes more significant, and therefore the z-score increases.

I think training LDpred2 on the Ferkingstad et al. (Nat Genet 2021) data (currently) has some challenges. First, these are BOLT-LMM summary statistics, which are different from standard least square...