Multiple-interval mapping for ordinal traits
Multiple-interval mapping for ordinal traits
复制标题
DOI：
10.1534/genetics.105.054619
复制
发表时间：
2006-07-01
期刊：
影响因子：
3.3
通讯作者：
Zeng, Zhao-Bang
中科院分区：
文献类型：
--
作者：
Li, Jian;Wang, Shengchu;Zeng, Zhao-Bang
已经开发了许多统计方法来绘制实验杂交群体中的多个数量性状基因座（QTL）。其中，多区间作图（MIM）可以同时定位具有上位性的QTL。然而，之前的MIM实现是针对连续分布的特征。在本研究中，我们基于阈值模型将 MIM 扩展到序数性状。该方法继承了MIM的特性和优点，可以拟合多重QTL效应和基础责任评分上位性的模型。我们研究了与该方法相关的许多统计问题，例如最大化和模型选择的效率和稳定性。我们还使用计算机模拟来研究该方法的性能并将其与其他替代方法进行比较。该方法已在 QTL Cartographer 中实现，以方便其在二元和序数性状的 QTL 作图数据分析中的通用应用。
Many statistical methods have been developed to map multiple quantitative trait loci (QTL) in experimental cross populations. Among these methods, multiple-interval mapping (MIM) can map QTL with epistasis simultaneously. However, the previous implementation of MIM is for continuously distributed traits. In this study we extend MIM to ordinal traits on the basis of a threshold model. The method inherits the properties and advantages of MIM and can fit a model of multiple QTL effects and epistasis on the underlying liability score. We study a number of statistical issues associated with the method, such as the efficiency and stability of maximization and model selection. We also use computer simulation to study the performance of the method and compare it to other alternative approaches. The method has been implemented in QTL Cartographer to facilitate its general usage for QTL mapping data analysis on binary and ordinal traits.