Selecting maximally informative sibships for OTL association analysis

Abstract

Fulker et al. (1999, have Am. J. Hum. Genet. 64, 259-267) have proposed a partitioning of association to between-sibship and within-sibship components and shown that a test based on the within-sibship variance is robust to population stratification. As in QTL linkage analysis, the efficiency of an association design may be improved by selecting phenotypically extreme sibships. We have developed a method of measuring the informativeness of sibships for QTL association analysis, in terms of the expected contributions of the sibship to the likelihood-ratio chi-square tests for between-sibship and within-sibship association. These expected contributions are calculated conditional on the measured trait values of the siblings, under a biometrical genetic model where QTL effects are parameterized by allele frequencies and genotypic means, and which allows for residual sibling correlations. This presentation will describe the method and its implementation in a freely-distributed computer program.