With the increasing accessibility of genome-wide data, there has been a corresponding surge of novel methods that leverage its high dimensionality. Typically, because approximately 1 million SNPs are examined individually, GWAS incur a high cost for multiple comparisons (p < 5E-8). In response, and consistent with the polygenic basis underlying behavioral traits (Plomin, Haworth, & Davis, 2009), recent research has begun to examine aggregate genetic influence and effects across genes and biological systems (Holmans, 2010; Neale & Sham, 2004; Purcell et al., 2009; Ramanan, Shen, Moore, & Saykin, 2012; Wang, Li, & Bucan, 2007; Wang, Li, & Hakonarson, 2010). Such gene- and system-level association analyses not only reduce the likelihood of false negatives by lowering the threshold for significance, but are also more compatible with the resolution (i.e., downstream neural, behavioral, and self-report measures) at which clinical, behavioral, and neural genetics research is conducted (Dudbridge, Gusnanto, & Koeleman, 2006; Nikolova, Ferrell, Manuck, & Hariri, 2011; Plomin et al., 2009).
