Knowing that the prior for a candidate gene selected on hypothesized biological rationale is low, there are a variety of other methods for selecting candidate genes for study that should produce more robust and reliable findings, including focusing on genes with either large main effects or stronger a priori evidence. For the latter, methods of gene selection that have a greater likelihood of producing meaningful cGxE results include focusing on candidates suggested by well-powered GWAS or meta-analysis, or by model organism work, ideally with replication. Identification of genotypes of substantial main effect is challenging and limited. For example, the ε2/ ε3/ ε4 polymorphism at the APOE locus is a known and important risk factor for Alzheimer’s and coronary heart disease (Corder et al., 1993; Ward et al., 2009). Such large-effect polymorphisms seem to be particularly compelling candidates for cGxE research because the genetic effect is already known; only an environmental modification of this effect is required for evidence for an interaction.
