Large-Effect cis-eQTLs in the Family Influence Essential Genes
It has been previously reported that cis-eQTLs based on population studies are depleted among essential genes.19 We hypothesized that if rare variation was indeed responsible for large-effect cis-eQTLs in the family, reduced impact of purifying selection on rare variants would result in family eQTLs disproportionately affecting essential genes. We tested this hypothesis in two ways: defining gene essentiality by (1) its degree of evolutionary constraint and (2) its centrality within a PPI network. To assess evolutionary constraint, we used dN/dS ratios between humans and chimps to compare large-effect cis-eQTL genes in the family to cis-eQTL genes in the population. We observed that large-effect cis-eQTL genes in the family had significantly higher conservation status than population cis-eQTL genes (Figure 3A). This was even more pronounced for genes with a rare and potentially regulatory variant within 5 kb of the TSS. By contrast, cis-eQTL genes in the population were less constrained for increasingly stringent p values.
We next applied PPI networks with the premise that genes that are more central in the network or have more connections to other genes are more essential than less connected ones. We found significantly higher connectivity for large-effect cis-eQTL genes in the family than for cis-eQTL genes in the population (Figure 3B). Furthermore, this contrast became stronger when we focused only on those genes that also contained a proximal rare and potentially regulatory variant (Figure 3B).