Data from: Mapping quantitative trait loci using selected breeding populations: a segregation distortion approach

Quantitative trait locus (QTL) mapping is often conducted in line crossing experiments where progeny are derived randomly from the original crosses. The detected QTL from such experiments are rarely relevant to breeding populations because they are not detected from the breeding populations. We developed generalized linear model methods to perform QTL mapping in directionally selected populations using a segregation distortion approach. A selected population is often small and thus has low power for QTL detection. The segregation distortion approach actually takes advantage of the small populations because small selected populations often reflected strong selection and thus possess a high degree of segregation distortion. We also developed methods to combine results of several populations and results from different types of data analyses from the same populations. Such a combined analysis can boost statistical powers. Simulation studies showed that the new methods of QTL mapping in selected populations are powerful. We illustrated the methods using two selected rice populations and detected several QTL responsible to yield selection. The new methods can be applied not only to rice breeding programs but also to breeding programs of all crops.数量性状位点（Quantitative trait locus, QTL）定位通常开展于杂交系实验中，此类实验的子代均由初始杂交组合随机产生。此类实验中检出的QTL极少能适配育种群体，因其并非从育种群体中直接检测得到。我们开发了基于分离畸变（segregation distortion）方法的广义线性模型（generalized linear model），用于定向选择群体的QTL定位。选择群体通常规模偏小，导致QTL检测的统计功效较低；而分离畸变方法实则可利用小群体的特性优势：经定向选择的小群体往往承载了较强的选择压力，因而呈现出高度显著的分离畸变。此外，我们还提出了整合多个群体的分析结果，以及同一群体内不同类型数据分析结果的方法，此类整合分析可有效提升统计功效。模拟研究结果显示，针对选择群体的新型QTL定位方法具备出色的检测效力。我们通过两个水稻选择群体对所提方法进行了实例验证，成功检出多个与产量选择相关的QTL。本方法不仅可应用于水稻育种项目，同样适用于所有作物的育种工作。