Second, in the absence of highly significant genetic associations for conduct disorder, and in view of the fact that it is a genetically complex trait, we encourage the use of polygenic scores in measured genetic studies of rGE and G × E (Wray et al., 2014). Polygenic approaches consider the weighted effects of SNPs across the genome, and thus characterize aggregate genetic risk in a way that is consistent with our understanding that many variants of small effect are likely to contribute to conduct disorder. A particular benefit of this approach is that it avoids the Type I error concerns associated with single variant/single gene cG × E approaches since the genetic effects are tested in aggregate. Typically, the results from large scale meta- or mega-analyses are used to create weighted linear combinations that reflect the degree to which an individual carries alleles that predispose him/her to conduct disorder. Thus, high polygenic scores indicate that an individual has a greater genetic predisposition to conduct disorder, and lower polygenic scores indicate that an individual has a lower genetic predisposition to conduct disorder. These scores can be carried forward into tests of rGE and G × E.
