Detecting genotype-environment interaction in monozygotic twin data: comparing the Jinks and Fulker test and a new test based on Marginal Maximum Likelihood estimation

Abstract

This article is concerned with the power to detect the presence of genotype by environment interaction (G x E) in the case that both genes and environment feature as latent (i.e., unmeasured) variables. The power of the test proposed by Jinks and Fulker (1970), which is based on regressing the absolute difference between the scores of monozygotic twins on the sums of these scores, is compared to the power of an alternative test, which is based on Marginal Maximum Likelihood (MML). Simulation studies showed that generally the power of the MML-based test was greater than the power of the Jinks and Fulker test in detecting linear and curvilinear G x E interaction, regardless of whether the distribution of the data deviated significantly from normality. However, after a normalizing transformation, the Jinks and Fulker test performed slightly better. Some possible future extensions of the MML-based test are briefly discussed.