Simulations indicate that this model is often considerably more powerful than the traditional approach that does not explicitly represent noncanonical genotypes (Supplementary Table 2 online). This joint approach is particularly powerful under certain scenarios, such as if a duplicated form of a ‘low-activity’ allele shows normal activity comparable with the wild-type allele (for example, when A and BB have similar phenotypes, different from B). In addition, the joint model can often disentangle SNP and CNP effects; for example, it can indicate whether a duplication or the allele that happens to be duplicated (or both) affects phenotype. The new release of the whole-genome association toolset PLINK19 now directly accepts Birdsuite output to perform these tests.
