A Score-Based Approach to Quantitative Trait Loci Mapping in Inbred Lines Using Flanking Markers
Kai Wang, Ph.D.
Program in Public Health Genetics
College of Public Health
University of Iowa
Friday, February 10, 2006, 12:00pm
1309 Health Sciences Learning Center (NOTE DIFFERENT LOCATION)
A score-based approach to quantitative trait loci (QTLs) mapping using flanking markers is proposed. This approach uses a parameterization under which the parameters of interest are locally orthogonal to nuisance parameters. By taking advantage of this property, explicit forms of score statistics are obtained for backcross and $F_2$ intercross under various analysis models. These score statistics are closely related to regression analysis. To overcome the problem that, at testing locations intermediate flanking markers, the variance of the
regression residual over-estimates the conditional phenotypic variance given QTL genotype,a consistent estimate of the latter is proposed. Simulation studies on $F_2$ intercross indicate that this score-based approach leads to a statistic more powerful than either the Haley-Knott regression method or the likelihood ratio test. This new approach is applied to an eQTL mapping study.