猪主要经济性状的基因组选择研究

A Study of Genome Selection Based on the Porcine Major Economic Traits

旨在系统比较GBLUP、SSGBLUP、BayesA、BayesB、BayesC、BayesLASSO、BSLMM和BayesR等8种方法对猪重要经济性状基因组选择的准确性。本研究以本实验室收集的2 585头大白猪达100kg日龄、达100kg背膘厚和母猪乳头数3个性状为分析对象,结合猪50K基因芯片分型数据,以加性模型为基础,利用5倍交叉验证比较8种方法的基因组选择准确性。研究发现,基因组选择的准确性与不同性状估计遗传力呈正相关。交叉验证结果表明,预测准确性最高的性状为达100kg日龄,但不同方法在不同性状中表现并不完全相同,在达100kg日龄和达100kg背膘厚中SSGBLUP基因组预测准确性均为最高,而在母猪乳头数中BayesA的基因组预测准确性最高。综上表明,对小样本开展基因组预测时,中、高等遗传力性状可以选择SSGBLUP方法,低等遗传力性状可以选择BayesA方法。如何优化和选择一种广泛适用于所有性状的方法,可能是未来研究的方向。

This study aimed to systematically compare the accuracy of genomic selection for important economic traits in pigs by using 8 models including GBLUP,SSGBLUP,BayesA,BayesB,BayesC,BayesLASSO,BSLMM and BayesR.The data of age at 100 kg,backfat thickness at 100 kg and teat numbers were collected from 2 585 Yorkshire sows and all the individuals were genotyped by using PorcineSNP50K Beadchip.The accuracy of genomic selection for the 8 models were compared by a 5-fold cross-validation procedure based on additive model.The results demonstrated that the accuracy of genomic selection were positively correlated with the calculated heritabilities of different traits.The cross-validation analysis indicated that the prediction accuracy of age at 100 kg was the highest among the 3 different traits,but different models performed dissimilarly in different traits.The prediction accuracy of SSGBLUP was the highest for both age at 100 kg and backfat thickness at 100 kg,and the prediction accuracy of BayesA was the highest for teat numbers.In conclusion,SSGBLUP model can be used for the traits with moderate and high heritabilities when conducting genomic prediction for small sample size and BayesA is suitable to the traits with low heritability.How to optimize and select a model that is applicable to all traits may be a research direction in the future.