Further support for region-specific cis-eQTL signals was provided by a comparison of our signals with those from other region-specific data sets. We separated our cis-eQTL signals into those with more evidence for region-specific effects (found in the tissue specified and significantly heterogeneous across tissues in our data set) and those with more evidence for pan-regional effects (found in the tissue specified but not significantly heterogeneous across tissues) (Supplementary Table 3). For pan-regional signals, we found roughly equal replication rates in each comparison data set when comparing the cerebellum to the frontal cortex, although there was still a tendency for replication rates to be higher when comparing same-tissue data sets. For region-specific signals, the difference in replication rates became much more pronounced, with performance noticeably better in the same-tissue comparisons. Finally, using monocytes as an outside comparison that may pick up pan-tissue eQTL signals, we noted that replication rates were noticeably better in our tissue-common signals (32.9%) than our tissue-complex ones (23.2%).
