Power of multifactor dimensionality reduction (MDR) for identifying gene-gene and gene-environment interactions

We have previously developed the multifactor dimensionality reduction (MDR) method to identify gene-gene and gene-environment interactions (Ritchie et al. AJHG 69, 2001). In brief, MDR is a method that reduces the dimensionality of multilocus information to identify polymorphisms associated with an increased risk of disease. This approach takes multilocus genotypes and develops a model for defining disease risk by pooling high-risk genotype combinations into one group and low-risk combinations into another group. Ten-fold cross validation and permutation testing are used to identify optimal models. The goal of this study was to evaluate the power of MDR for identifying gene-gene and gene-environment interactions in the presence of common sources of noise. Using four different epistasis models, we simulated discordant sib-pairs with 5% genotyping error, 5% phenocopy. 20% phenocopy, or 50% genetic heterogeneity. MDR was able to identify the functional loci with 80-98% power in the presence of genotyping error or phenocopy, and 47-78% power in the presence of genetic heterogeneity. These results demonstrate that MDR is a powerful method for identifying and characterizing gene-gene and gene-environment interactions, even in the presence of some common sources of noise.