文档下载 文档下载导航
设为首页 | 加入收藏
搜索 请输入内容:  
 导航当前位置: 文档下载 > 所有分类 > Markov Chain Monte Carlo Based Inference for Quantitative Trait Loci in Plant Breeding
侵权投诉

Markov Chain Monte Carlo Based Inference for Quantitative Trait Loci in Plant Breeding

Markov chain Monte Carlo based inference for quantitative trait loci in plant breeding

group is used as a test bed to demonstrate the use of the Markov chain Monte Carlo method as an alternative to interval mapping or regression methods in common use today. Statistical methods for estimating single quantitative trait locus (QTL) in a linkage group are fairly well understood when su cient assumptions are made. For example, the interval mapping LOD score approach (Lander and Botstein 1989) uses the EM method to pro le the locus across the linkage map. Recently, this idea has been extended to estimating several quantitative trait loci (QTLs) using likelihood and/or regression methods (Zeng 1994; Jansen and Stam 1994; Churchill and Doerge 1994a). However, these approaches become much more complicated as larger models are considered. Markov chain Monte Carlo (MCMC) methods promise to extend the domain of such inference by weakening assumptions, such as allowing multiple QTLs and relaxing the normal distribution of traits. MCMC methods, particularly the Metropolis-Hastings algorithm and Gibbs sampling, have recently been applied to several genetic problems including ordering of markers o

n linkage maps (Stephens and Smith 1993) and linkage analysis in complex pedigrees (Thompson 1991). Markov chain Monte Carlo ideas have been developed in recent articles by Tierney (1991), Besag and Green (1993) and Smith and Roberts (1993). Geyer (1992) discusses practical considerations to reduce bias and autocorrelation and to assess error variation using time series methods. In this paper we develop a MCMC approach for estimating a single QTL in one linkage group for a simple breeding system by augmenting the marker and trait data with the missing QTL genotypes at a putative QTL. While this problem can be readily handled by pro le likelihood computations, it demonstrates the power of these methods. Analagous MCMC methods are proposed for polygene (multiple QTL) inference in more complicated settings. Section 2 outlines a plant breeding scheme which allows us to only consider haploid type. Section 3 presents a model for a single QTL, detailing the relationship between trait data, marker data on the linkage map, the missing genotypes and model parameters. The interval mapping method is brie y outlined at the end of this section. Section 4 develops the MCMC method for a single QTL. Section 5 proposes extensions of the MCMC method to polygenes and other potential generalizations. Finally, simulations in Section 6 examine the behavior of the MCMC method relative to interval mapping.

2

上一页第3页

你可能喜欢

热门文档

相关文档

站点地图 | 文档上传 | 侵权投诉 | 手机版
本站所有资源均来自互联网,本站只负责收集和整理,均不承担任何法律责任,如有侵权等其它行为请联系我们.
文档下载 Copyright 2013 doc.guandang.net All Rights Reserved.  -  email