A primary motivation for our study was to investigate eQTLs as plausible modes of action for association hits from GWAS. To this purpose, we looked for overlap of our cis-eQTL signals with known GWAS association signals (Table 1 and Supplementary Table 7). The US National Human Genome Research Institute (NHGRI) GWAS catalog (accessed 7 March 2014), when restricted to SNPs in our 1000 Genomes imputed SNP data set, revealed 385 genome-wide significant SNPs (P < 5 × 10−8) associated with brain-related traits, such as multiple sclerosis, bipolar disorder and Parkinson’s disease, and 3,949 SNPs associated with other traits, such as height, type 2 diabetes and Crohn’s disease. Of these, 17.4% and 20.8% respectively matched cis-eQTL signals in our data set. When we restricted the analysis to SNPs associated specifically with adult-onset neurological disorders, we found that 23.1% of risk SNPs matched cis-eQTL signals. A substantial percentage of brain-related GWAS-eQTL signals acted at a distance, including many cases where the GWAS-signal eQTL signal set (GWAS-eQTL plus all markers with r2 >0.8) resided in a different gene or even a different associated
