We performed fine-mapping based on summary statistics and matched LD matrix for each causal block of each trait using three commonly used tools, namely FINEMAP (27), PAINTOR (26) and CAVIARBF (28) (Supplementary Table S2). We assumed that there was only one causal variant in a causal block and used the recommended parameters of the tools. These fine-mapping tools can report the posterior probability (PP) of each variant being causal in the specific model. A credible set is the set of variants with a sum of PP of more than α, which means considering the cumulative sum of PPs from the largest to smallest until it is not smaller than α. In CAUSALdb, we reported potential causal variants within the credible set upon the adjustment of α. The code for reproducing the CAUSALdb GWAS fine-mapping procedure can be found at https://github.com/mulinlab/CAUSALdb-finemapping-pip.
