To identify potential causal variants in GWAS signals, we fine-mapped each genome-wide significant (p < 5e-8) region from the 2,444 GWAS endpoints. Each region was fine-mapped with SuSiE and FINEMAP. We used in-sample LD for fine-mapping.

We used a 3-megabase window (+- 1.5M) around each lead variant, merged overlapping regions into one, and used these regions for fine-mapping.

Edit on GitHub