Analysis of RNA expression by RT-qPCR produced two unexpected observations. First, although the rs279858*C-allele was enriched in low expression Cluster 1 cell lines (Table 1), the within-cluster expression of chr4p12 GABAA genes did not differ by GABRA2 genotype (Figure 5). Therefore, the genotypic association with expression level illustrated in Figure 1A reflects the greater frequency of rs279858*C-allele carriers in Cluster 1 vs. Cluster 2 (0.94 vs. 0.47), suggesting that the genetic effect linked to rs279858 strongly determines Cluster membership. The absence of an rs279858 genotype effect within Cluster 2 cell lines suggests that a simple Mendelian genetic model is not sufficient. One model consistent with our results includes a cis-acting genetic locus (in linkage disequilibrium with rs279858) that influences the expression of the chr4p12 GABAA gene cluster together with a recently discovered phenomenon of random clonal allelic-biased expression of a subset of autosomal non-imprinted genes (Chess, 2012). Specifically, we hypothesize that a random stochastic process occurring during the reprogramming of individual fibroblasts to generate iPSC clones results in random allelic bias for the expression of chr4p12 GABAA genes. In this
