基于最小二乘估计的多性状QTL复合区间定位法(英文)

本文给出了多性状数量性状基因座复合区间定位法（ＭＣＩＭ）的一种改进方法．它是基于多重多元回归模型的，因此可以用最小二乘法来配合模型．使得它在计算上要比原来基于最大似然法的ＭＣＩＭ简单快捷得多．给出了多重逐步回归的方法，可用来筛选能提供遗传信息的标记作为ＭＣＩＭ模型中的余因子．文中还建议用排列测验的方法估计所需的显著阈值．给出了一个实际例子．

Multiple trait composite interval mapping of QTLs based on least squares estimation

A modified approach of multiple trait composite interval mapping (MCIM) of quantitative trait loci (QTLs) is proposed in this paper. The modified MCIM is based on a multivariate multiple regression model which can be fitted easily with the method of least squares estimation. This makes its computation much simpler and quicker than the original maximum likelihood based MCIM. An approach of multivariate stepwise regression is proposed to screen for informative markers as cofactors for MCIM. It is also suggested that the method of permutation test be used to estimate an appropriate significance threshold. A practical example is given.