In accordance with previous eQTL studies, we observe a symmetric distribution of eQTLs around the transcription start site (TSS), with the strongest, most highly replicated, signals directly at the TSS. The simplest explanation for the relationship between distance to TSS and replication across populations is one of statistical power, as the long-distance smaller effects are less likely to be replicated. However we cannot exclude the possibility that these patterns, at least in part, reflect true biology such as the nature of long-distance enhancers. We applied factor analysis to the gene expression data to account for global non-genetic effects on the expression profiles and increased our power to detect eQTLs, especially those of smaller effect. The additional cis-eQTL associations detected using the residual gene expression profiles (‘REDUCED’ data) have very similar characteristics to those detected using the straightforward normalized and PCA-corrected data, however the degree of across-population replication was higher for the ‘REDUCED’ data, which provides confidence that the method is not simply adding false positive associations. Indeed, this demonstrates that by applying dimension reduction, we do not introduce bias, rather
