Our HLC cohort did successfully identify a large number of eGenes shared with the GTEx liver cohort, as well as a large number of eGenes that did not independently qualify as eGenes in the GTEx liver cohort, including genes expressed only in hepatocytes (such as ANGPTL3). We were able to use these findings to discover and validate a number of functional genes—HLC eGenes CPNE1, VKORC1, and UBE2L3 and HLC-specific differential ASE gene ACAA2 modulated blood lipid levels in mouse models—as well as functional variants. One possible reason why many HLC eGenes were not independently observed in the GTEx liver cohort (at FDR < 5%) is a limitation in power offered by the currently available GTEx cohort. Indeed, observations from much larger primary human liver eQTL studies (as large as ~1,000 samples) provide evidence for CPNE1, ANGPTL3, and FRK being liver eGenes (Teslovich et al., 2010; Innocenti et al., 2011). Thus, we expect that comparisons of much larger, better-powered HLC cohorts and liver cohorts would reveal substantially more overlap of eGenes than we observed in the present study. The establishment of
