Methodology of Mapping Quantitative Trait Loci for Binary Traits in a Half-sib Design Using Maximum Likelihood

Yin Zongjun,Wang Chunkao,Zhang Jigang,Zhang Qin,Ding Xiangdong

doi:10.5713/ajas.2005.1669

Abstract

Maximum likelihood methodology was applied to analyze the efficiency and statistical power of interval mapping by using a threshold model. The factors that affect QTL detection efficiency (e.g. QTL effect, heritability and incidence of categories) were simulated in our study. Daughter design with multiple families was applied, and the size of segregating population is 500. The results showed that the threshold model has a great advantage in parameters estimation and power of QTL mapping, and has nice efficiency and accuracy for discrete traits. In addition, the accuracy and power of QTL mapping depended on the effect of putative quantitative trait loci, the value of heritability and incidence directly. With the increase of QTL effect, heritability and incidence of categories, the accuracy and power of QTL mapping improved correspondingly.

Highlights

Since the advent of molecular markers, novel statistical techniques to detect and map individual genes affecting quantitative traits have been developed and subsequently refined
The results showed that the threshold model has a great advantage in parameters estimation and power of quantitative trait loci (QTL) mapping, and has nice efficiency and accuracy for discrete traits
Most methods of QTL mapping are based on the interval mapping method, using either least squares regression or maximum likelihood (Haley et al, 1992; Grignola et al, 1996; Weller, 2001; Feingold et al, 2002; Liu et al, 2004; Tae-Hun Kim et al, 2004), in which a putative QTL is fitted at every possible position along a chromosome and the best fitting position is taken as the estimate of QTL position

Summary

Introduction

Since the advent of molecular markers, novel statistical techniques to detect and map individual genes affecting quantitative traits have been developed and subsequently refined. In maximum likelihood (ML), the likelihood function of the observed data is maximized with respect to all parameters in the model postulated, and hypothesis testing is based on the likelihood ratio statistic. Examples are mastitis, calving ease and insemination success or failure in cattle. Such traits vary in a discontinuous manner but are not inherited in a simple Mendelian fashion. Genetic analyses for categorical traits are difficult because the observed phenotype cannot be described by a linear function of genetic and environmental effect. For such traits threshold models are more appropriate than linear models, which assumes that there is a continuous underlying variable which determines the expression of discrete trait.

Methods

Results

Conclusion