There is empirical evidence that eQTLs are over-represented in GWAS signals, and several reports have linked AD association signals with eQTLs [31, 38, 39, 79, 80]. Our study adds to this growing body of evidence, as we find that AD GWAS signals are significantly over-represented in our eQTLs, thus strengthening the significance of these genetic associations and providing potentially causal mechanisms of action for our eQTL findings. For example, in addition to NNAT and PSMB5, eQTLs for a long non-coding RNA (PKI55; rs13392737), adaptor related protein complex 1, sigma 1 subunit (AP1S1; rs10279545) and translocation associated membrane protein 1 (TRAM1; rs2959574) were also associated with AD at p = p = 9.1x10-4 and p = 6.2x10-3, respectively. In particular, NNAT and AP1S1 have prior evidence of involvement with AD [81]. NNAT, an imprinted gene expressed early in brain development, was shown to regulate dendritic calcium levels in hippocampal neurons and was differentially expressed in the NAc, PFC, and VTA of an acute ethanol response mouse model [82, 83]. AP1S1, a clathrin-related protein involved in membrane trafficking and endocytosis, as well
