The heritability of a quantitative trait (y) can be represeented as the linear regression of offspring vlaues (yi) on mid-parental values (ymp), such that yi = Bo,mp ymp + e. Where B, the slope of the regression of yi on ymp, is taken to be an estimate of the heritability of the trait (y), and e is a normally distributed residual effect. The inclusion of the effect of a marker g(L), in the regression can provide a test of associdation between the trait (y) and the marker, e.g., yi = Br ymp +g(L) + e. The difference between the slopes of these two regression Bo,mp - Br, is the heritability attributable to the marker L. This approach is a fundemental extension to quantitative genetic theory and may prove useful for the detection of small effects due to functional single nucleotide polymorphims (SNPs) in the determination of whether that SNP is responsible, at least in part, for phenotypic variation for a quanitative trait. The first use of this method (ROMP) was published during the past year, and the theoretical basis of the method has been submitted. Two other manuscripts using ROMP are in preparation.
