To further prioritize likely causal genes and variants, we performed co-localization analyses integrating fine-mapped GWAS results, using the CAUSALdb pipeline (Methods), and eQTL data from a meta-analysis of three brain datasets54 applying a fixed-effect model. First, we adopted the Coloc method55,56, which revealed 13 genes with strong evidence for both GWAS-association, eQTL-association and co-localization (i.e., with a posterior probability of PPH4>0.8; Supplementary Table S10). The three top-ranked genes were: FURIN, NEGR1 and CKS2. Secondly, we conducted the eQTL and GWAS CAusal Variants Identification in Associated Regions (eCAVIAR) approach57, in which both eQTL and GWAS were fine-mapped and the product of posterior probability (CLPP) was calculated, prioritizing variants with at least a single variant with CLPP >= 0.01. The eCAVIAR approach revealed five prioritized variants located in three genes: FURIN, GPR27 and TCTA (Supplementary Table S11 and Figure S11).
