We also identified genes that exhibited larger ASE effects in the family than in the population. We found that 223 of the 1,777 genes we tested had larger ASE effect sizes in the family (CI > 0.95, FDR = 1,777 × 0.05 / 223 = 40%; Figure 1B; Figure S25). We expected that on an individual basis, the family and population would actually have the same distribution of ASE effect sizes (no excess of large effects, FDR = 1). We controlled for some initially observed excess in the family by matching read depths via downsampling; however, this did not address all the excess in the family, and unknown factors still remained. We expected that any excess, however, would only add noise to subsequent rare-variant enrichment analyses, and we further validated large ASE effects by using evidence from IBD siblings (Figure S25). In addition, we applied ASE to support discoveries of cis-eQTLs in the family; by stratifying their degree of effect size relative to those in the population, we detected a proportionally increased enrichment of detectable ASE (significant ASE sites defined as allelic imbalance > 0.05, binomial test p value < 0.05; Figure S21). This relationship supports a potential regulatory role of rare variants because it indicates that large-effect cis-eQTLs in the family might be the consequence of heterozygous variants that manifest in ASE. This idea is further supported by our observation of a direct and simple linear relationship between cis-eQTL effect size among children and ASE effect size among parents (Figure S20).