As many eQTLs are shared among tissues9,25,26, we sought to place iPSC eQTLs in the broader context of somatic tissues. We assessed iPSC eQTL replication across 44 GTEx tissues (lead eQTLs and proxy variants, r2 > 0.8, defining replication as P < 0.01/45; Methods), revealing 2,131 eQTLs that were specific to iPSCs (Fig. 4a). We also considered secondary eQTLs, identifying a similar proportion of iPSC-specific genetic effects (both 32%). Most tissue-specific signals (72%) occurred in genes with at least one GTEx eQTL that was not in high linkage disequilibrium (LD) with the lead iPSC eQTL variant, suggesting that iPSC-specific eQTLs are frequently driven by alternative regulatory variants. Only 11% of the iPSC-specific eQTLs could be attributed to tissue-specific gene expression (Fig. 4b), despite greater numbers of expressed genes in iPSCs compared with somatic tissues (Extended Data Fig. 7). Similarly, most somatic tissue-specific eQTLs were also driven by alternative regulatory variants, with only testis showing a substantial fraction (16%) of eQTLs attributable to tissue-specific gene expression (Fig. 4b). Using alternative methods for eQTL detection and assessing the extent of sharing between
