Trans-eQTL analysis was also very sensitive to statistical power. As other investigators have reported previously, trans effects tend to be weaker and replicate less often.14,16 From a biological standpoint, given the large genomic distance between causal locus and affected gene, trans-eQTL more than likely influence expression through an intermediary such as a transcription factor or a microRNA regulator. Under these circumstances, any other factors that partake in this regulatory process could dilute the initial signal by adding noise. For instance, trans-eQTL analysis of IBD-associated SNPs resulted in many associations with P values as low as 10−5 and 10−6; however, none of these survived correction for more than 1.4 million total comparisons given the sample size limitations of this study.
