Testing the robustness of the likelihood-ratio test in a variance-component quantitative-trait loci-mapping procedure

Detection of linkage to genes for quantitative traits remains a challenging task. Recently, variance components (VC) techniques have emerged as among the more powerful of available methods. As often implemented, such techniqu es require assumptions about the phenotypic distribution. Usually, multivar iate normality is assumed. However, several factors may lead to markedly no nnormal phenotypic data, including (a) the presence of a major gene (not ne cessarily linked to the markers under study), (b) some types of gene x envi ronment interaction, (c) use of a dichotomous phenotype (i.e., affected vs, unaffected), (d) nonnormality of the population within-genotype (residual) distribution, and (e) selective (extreme) sampling. Using simulation, we h ave investigated, for sib-pair studies, the robustness of the likelihood-ra tio test for a VC quantitative-trait locus-detection procedure to violation s of normality that are due to these factors. Results showed (d) that some types of nonnormality, such as leptokurtosis, produced type I error rates i n excess of the nominal, or alpha, levels whereas others did not; and (b) t hat the degree of type I error-rate inflation appears to be directly relate d to the residual sibling correlation. Potential solutions to this problem are discussed. Investigators contemplating use of this VC procedure are enc ouraged to provide evidence that their trait data are normally distributed, to employ a procedure that allows for nonnormal data, or to consider imple mentation of permutation tests.