We investigated the enrichment of cis-eQTL signals with respect to sentinel marker type and genic location. Sentinels do not necessarily represent true causal variants, and this introduces noise into our analyses. However, to the extent that we do find some significant relationships, these can only arise from there being some overlap of true causal variants with sentinels. Furthermore, we imputed to the 1000 Genomes data set before performing eQTL analyses, which increases the likelihood of overlap between sentinels and true causal variants.
