and on significant regional heterogeneity using our modified test of heterogeneity). One of the most interesting examples of such markers is rs73009150, which was a cis-eQTL signal for RNF214 in medulla, for PAFAH1B2 in cerebellum and putamen, and for SIK3 in putamen (Fig. 4a-c). We note that one explanation for these results could be the existence of multiple functional variants in linkage disequilibrium. Nevertheless, taken together with the regional heterogeneity of cis-eQTL signals, these findings highlight the potential dangers of extrapolating eQTL findings from one tissue type to another, even if they are highly related.
