We sought to identify genetic influences on adolescent BD through a multifaceted approach. We initially characterized results from a standard GWAS by estimating the variance explained by common SNPs, and used gene- and pathway-based tests to identify potential novel candidate genes and pathways. Results from the estimation of sample variance explained by all genotyped SNPs and significant gene-based tests suggest there is a real genetic signal to be detected within the noise. However, the current sample is likely underpowered to detect realistic effect sizes of individual SNPs. Further, the lack of correspondence between pathway analyses in the CADD and replication samples may be due to limited power, or qualitative differences in the genetic effects on BD across different ages (adolescent versus adult) or sampling distributions (over-sampled for BD versus community-representative). Key to the search for causal genetic pathways underlying BD will be the availability of increasingly large, thoroughly phenotyped samples. Although the current analyses did not identify specific loci associated with BD, we demonstrate substantial heritability due to effects of common SNPs. Larger studies with appropriate phenotypes could well allow successful identification of common variants associated with BD.
