association (no shift in P value distribution) with bipolar disorder or type 2 diabetes among the blood eQTLs (P > 0.05) (Fig. 8A). This tissue specificity of autoimmune enrichment results suggests they are not due to the confounding effects that result from similar underlying genomic properties between eQTL and disease association regions. More generally, we observed trait-specific levels of enrichment for the WTCCC disease traits among the nine different single-tissue eQTL sets.
