Here is presented an approach to testing whether the effect of a candidate
gene on a quantitative trait is dominant and for testing whether the effect
is recessive, The approach uses parental genotype information in nuclear f
amilies to adjust for bias due to population admixture, The approach is app
licable regardless of the nature of the sampling, The results of an applica
tion of the methods to a candidate mutation for diabetic nephropathy are us
ed for illustration. Copyright (C) 2000 S. Karger AG, Basel.