Testing for equality of standardized composite measures of linkage disequilibrium.

We described the approximate variance for a composite measure of linkage disequilibrium (LD), which depends only on genotype frequencies and allows for departure from Hardy-Weinberg disequilibrium (Hamilton et al. 2006). Here we extend this work to allow for inferences about differences in composite LD coefficients, as we illustrate with two examples. Simulations demonstrate consistency of type I error rates under most conditions. Calculations for power and sample size are also given.