We used our regulatory network based on Hi-C, QTLs, and activity relationships to connect non-coding GWAS loci to potential disease genes. In particular, for the 142 SCZ GWAS loci, we identified a set of 1111 putative SCZ-associated genes, covering 119 loci (the SCZ genes) (Fig. 6A) (47). Of these, 321 constitute a “high-confidence” set supported by more than two evidence sources (e.g., QTLs and Hi-C) (Fig. 6, A and B, and fig. S45); examples include the CHRNA2 and CACNA1C genes (Fig. 6, B and C). Overall, the SCZ genes represent an increase from the 22 genes reported in an earlier QTL study and a larger number than can be linked simply by genomic proximity (176) (Fig. 6A) (11, 47). The majority of SCZ genes were not even in LD with the index SNPs (~67%, or 748 of 1111 genes with r2 < 0.6) (fig. S45), consistent with the fact that regulatory relationships often do not follow linear genome organization (13).
