ASE analysis can also be used as orthogonal confirmation of eQTL effects, on the basis of the expectation that individuals who are heterozygous for a cis-eQTL variant should manifest biased allelic ratios in the eQTL target gene. We performed this analysis at the genome-wide level across all the tissues in the GTEx data set, including the tissues with sample sizes that were too small for eQTL analysis. Examining the significant eQTLs identified by single-tissue analysis for the nine main tissues, we looked at their ASE effects separately in each of the 42 tissues, calculating the odds ratio of significant versus nonsignificant ASE for eQTL heterozygote versus homozygote individuals (e.g., for thyroid eQTLs we looked for ASE in all 42 tissues, then for blood eQTLs and so on) (figs. S20 and S21). In addition to replication of eQTL signals in the discovery tissue, we can estimate how relevant our eQTL findings from nine tissues are to a wide variety of other tissues (independent of sample size or allele frequency) and then assess which tissue is the best proxy for capturing regulatory effects in another tissues of interest.
