showing differential expression (Table S10). Of these 70 transcripts, 14 were within one of the BD-associated GWAS loci, including both ANK3 and CACNB3, which replicates with digital mRNA counting assay the qRT-PCR results and Western blot results shown in Figure 3. These 14 transcripts correspond to a frequency 78% of the 18 BD candidate genes assayed. Of the other non-BD candidate genes affected in the probe set, there were 56 genes corresponding to a frequency 49.6% of the total of the 113 non-BD candidate gene. Comparing these frequencies (77.8% BD genes vs. 49.6% non-BD genes) represents a statistically significant greater enrichment of affected BD candidate genes than that expected by chance (p< 0.05; two-tailed, Fisher’s exact test) (Figure S5). In support of the notion that miR-34a targets a molecular network of interrelated proteins critical to neurodevelopment, analysis of protein-protein interactions amongst these newly validated (ANK3, CACNB3, DDN), previously validated (SYN1) targets and genes implicated by neuropsychiatric genetics16, 19, 23 revealed a highly connected interaction network (Figures 4B and S6; See Supplemental text for details). Together, these findings support the hypothesis that elevation of miR-34a levels coordinately affects functionally connected pathways important for brain development and that dysregulation of miR-34a may
