家畜全基因组选择中的单步贝叶斯方法  

Single-step Bayesian Method for Genome-wide Selection in Domestic Animals

作　　者：梅步俊[1,2] 王志华[3] 

机构地区：[1]河套学院农学系,古巴彦淖尔015000 [2]爱荷华州立大学动物科学系,爱荷华州埃姆斯50010 [3]河套学院土木工程系,古巴彦淖尔015000

出　　处：《基因组学与应用生物学》2016年第10期2668-2675,共8页Genomics and Applied Biology

基　　金：国家自然科学基金(项目批准号:31460594);国家留学基金委项目(项目批准号:201308155140);河套学院教学研究项目(项目批准号:HTXYJZ14005);内蒙古自治区高等学校科学研究项目(项目编号:NJZY262)共同资助

摘　　要：随着标记信息可以被越来越多的应用在家畜育种中,许多基因组选择(GS)方法使得育种工作者可以利用家畜早期的基因型数据提前对其进行选择。结合系谱、表型和基因型数据,我们可以缩短家畜的世代间隔,提高家畜遗传价值估计的准确性,进而加速其遗传改良速度。近年来,和广泛使用的多步基因组选择策略相比,业界更推崇基于在系谱关系矩阵中增加基因组信息的单步遗传评估方法。即使通常的基因组选择方法依然是多步方法,如GBLUP法,但是基于单步基因组模型进行的基因组评估能提供更为准确的结果。本研究的目的是引入单步贝叶斯方法,此方法可以用贝叶斯回归模型直接计算单核苷酸多态性(SNP)的效应,同时我们使用模拟方法评估模型的性能。研究结果显示:QTL数目对单步贝叶斯方法的准确性无影响,但其准确性受遗传力的影响。同时,其准确性随着测序个体数的增加而增加。我们也讨论了与使用单步贝叶斯方法相关的问题,并详细描述了一些与之有关的统计理论和算法问题。With the growing availability of marker information in animal breeding, most genomic selection （GS） methods can permit breeders to preselect livestock based on genotypes early in life. The union of pedigree, phenotypic and genotypic information can accelerate genetic progress by shrinking the generational interval and raising the accuracy of genetic merit estimates. Recently, compared with the universally used multi-step genomic strategy for genomic selection, the more single-step genetic evaluation is highly proposed based on the pedigree relationship matrix augmented with genomic information. Even though most common approaches for genomic selection are still based on multiple-step method, like GBLUP, the single-step genomic models can provide more accurate genomic evaluation. The objectives of this study were to introduce a single-step bayesian genomic approach, allowing a direct calculation of single nucleotide polymorphism （SNP） effects by Bayesian regression models, and to evaluate model performance using simulation method. The results show the number of QTL has little effect on accuracy of single-step bayesian method, and its accuracy is sensitive to heritability. Meanwhile, the accuracy of new method increases with number of genotyped individuals. We discussed various issues related to implementation of the single-step bayesian method, and described several details of statistical theories and computing algorithms.

关 键 词：基因组评定 基因组选择 单步贝叶斯方法 动物育种 计算生物学 

分 类 号：S813.1[农业科学—畜牧学]