贝叶斯回归分析方法及其在QTL作图中的应用

以实例详述了运用贝叶斯统计进行线性回归分析的原理和方法，并将其推广应用于F2群体的QTL作图。该方法可同时实对QTL基因型、位置、效应及模型剩余方差等参数的贝叶斯估计，其可行性和有效性通过计算机模拟数据进行了分析验证。结果表明：尽管随着遗传力的提高，QTL位置和效应估计值的准确度随之提高，但在设定的3种遗传力(5％、10％和20％)下相差不大，均很准确；而估计值的精确度则随着遗传力的提高而明显提高。因此，在QTL遗传力较低时，为了提高QTL位置和效应估计的准确度与精确度，适当的样本容量是必要的。

Bayesian method of linear regression analysis and its application in QTL mapping

The theory and method of linear regression analysis by using the Bayesian statistics were introduced in detail and the analyzing procedure was demonstrated with a typical real example. Furthermore, it was extended and applied to QTL analysis in F_2 mapping population. The method may implement the Bayesian estimation of QTL genotypes, position, effects and model residual variance simultaneously, and its feasibility and validity were analyzed with simulation data. The results showed that although the accuracy of estimates of QTL position and effects would enhance as the increasing of QTL heritability, the differences were very small among different treatments including 5%, 10%, and 20% of QTL heritability. However, the precision of estimates would enhance significantly as the increasing of QTL heritability. This result indicates that the appropriate sample size will be necessary for QTL mapping in low QTL heritability to improve the precision of estimates of QTL position and effects.