The residualized expression data were nearly perfectly normally distributed, and so the bivariate normal model of Wright 85 for sib-pair IBD mapping was applied to the DZ pairs, offering a potential improvement over the Haseman-Elston approach.8 MERLIN 86 was run on the thinned set of markers used for stratification analysis, and probabilistic IBD estimates produced at each marker closest to or within each gene. A full maximum likelihood approach was applied for an additive model for the effect of each increment of IBD on DZ twin correlation as a function of IBD status, thus extracting maximum information, and converted to local h2-equivalents as the proportion of variation in the trait explained by local IBD status.
