QTL mapping in multiple families using logistic regression

Abstract

This study compares logistic regression (LR) with maximum likelihood (ML) methods for mapping quantitative trait loci (QTL) in multiple half-sib families under selective or full genotyping strategies, with various levels of marker informativeness and marker interval. In ideal conditions involving evenly located polymorphic markers and all individuals genotyped, both LR and ML methods showed a high power of detecting QTL and produced accurate estimates of QTL locations and effects. Under selective genotyping strategy, the power of ML is limited by regions with low information content. The LR method performed better than ML and is a straight-forward and robust method for this case.