Our data show clearly that patterns of expression are measurably different across brain tissues. QTL analyses reveal an abundance of eQTLs and, as previously reported, significant eQTLs are predominantly cis in nature [2], [5]. It is notable, particularly given the systematic differences in expression patterns among the four tissues, that the majority of large effect eQTLs detected were consistent across brain tissues. Large effect tissue-specific QTLs are also observable in the current data set, suggesting that there are some genetic effects on expression that are dependent on the tissue type used irrespective of expression levels of the mRNA. Our results that reached statistical significance are consistent with recent work showing that many eQTLs are likely tissue specific, but they typically have a smaller effect and are more distant [23]–[25]. However within our data when we observe a QTL that is only significant in a single tissue, examination of the correlation between the same SNP and trait in other tissues generally reveals a suggestive, although non-significant, correlation in the other tissues; thus both large and many moderate size effects appear to
