中文摘要：

基因组育种值估计是基因组选择的重要环节，基因组育种值的准确性是基因组选择成功应用的关键，而其准确性在很大程度上取决于估计方法。目前研究和应用最多的基因组育种值估计方法是贝叶斯（Bayes）和最佳线性无偏预测（BLUP）两大类方法。文章系统介绍了目前已提出的各种Bayes方法，并总结了该类方法的估计效果和各方面的改进。模拟数据和实际数据研究结果都表明，Bayes类方法估计基因组育种值的准确性优于BLUP类方法，特别对于存在较大效应QTL的性状其优势更qt显。由于Bayes方法的理论和计算过程相对复杂，目前其在实际育种中的运用不如BLUP类方法普遍，但随着快速算法的开发和计算机硬件的改进，计算问题有望得到解决；另外，随着对基因组和性状遗传结构研究的深入开展，能为Bayes方法提供更为准确的先验信息，从而使Bayes方法估计基因组育种值准确性的优势更加突出，应用将会更加广泛。

英文摘要：

Estimation of genomic breeding values is the key step in genomic selection. The successful application of genomic selection depends on the accuracy of genomic estimated breeding values, which is mostly determined by the estimation method. Bayes-type and BLUP-type methods are the two main methods which have beenwidely studied and used. Here, we systematically introduce the currently proposed Bayesian methods, and summarize their effectiveness and improvements. Results from both simulated and real data showed that the accuracies of Bayesian methods are higher than those of BLUP methods, especially for the traits which are in- fluenced by QTL with large effect. Because the theories and computation of Bayesian methods are relatively complicated, their use in practical breeding is less common than BLUP methods. However, with the develop- ment of fast algorithms and the improvement of computer hardware, the computational problem of Bayesian methods is expected to be solved. In addition, further studies on the genetic architecture of traits will provide Bayesian methods more accurate prior information, which will make their advantage in accuracy of genomie estimated breeding values more prominent. Therefore, the application of Bayesian methods will be more exten- sive.