基于GBLUP与惩罚类回归方法的猪血液性状基因组选择研究

A Study of Genomic Selection on Porcine Hematological Traits Using GBLUP and Penalized Regression Methods

旨在探讨GBLUP与惩罚类回归方法用于猪血液性状基因组选择的相关问题。以本实验室收集的免疫资源猪群体13个血液性状为分析对象,结合Illumina公司猪SNP60K基因芯片分型数据,以加性模型和加性-显性模型为基础,利用GBLUP和3种惩罚类回归方法(ridge、lasso与elastic-net)开展基因组选择分析。研究发现,基因组选择的准确性与性状芯片遗传力估计值呈正相关。交叉验证分析结果表明,4种方法对13个血液性状预测准确性最高的性状均是MCV(平均红细胞体积),而加性模型和加性-显性模型的预测准确性在不同性状中的表现不同。在多数性状中,lasso和elastic-net回归的预测准确性低于ridge回归和GBLUP法,但在NE%(嗜中性细胞百分比)等少数性状中则刚好相反。综上说明,没有适用于所有性状的最佳基因组预测方法,基因组预测方法的选择应考虑目标性状的遗传特性。本研究为猪免疫性状基因组选择的实际应用提供了重要参考信息。

This study aimed to explore the application of GBLUP and penalized regression methods in genomic selection of the hematological traits in pigs. We chose 13 hematological traits from the immune resource population collected by our laboratory as the analyzed objects. We used the genotyping data of Illumina PorcineSNP60 Genotyping Beadchip to conduct the genomic selection analysis, in which GBLUP and 3 penalized regression methods （ridge, lasso and elastic-net） were used based on additive model and additive-dominance model. The results showed that the accuracy of genomic selection was positively correlated with estimated values of chip heritabilities of traits. The results of cross-validation analysis showed that the MCV （mean corpuscular volume） had the highest prediction accuracy among 13 hematological traits. The prediction accuracy of additive model and additive-dominance model were different in different traits. In total trend, the prediction accuracy of the lasso and elastic-net regressions were lower than that of the ridge regression and GBLUP. But in a few traits, such as NE%, it was opposite. In conclusion, there is no optimal genomic prediction method that could be suitable for all traits, and we should consider the genetic characteristics of the target traits when choosing a genome evaluation method. This research provides important reference information for the practical application of genomic selection for immune traits in pigs.