摘要
在基因组选择中(GS),相比单性状模型,多性状模型有诸多优势:可以利用到性状间的遗传相关信息,提高预测可靠性;当直接选择主要性状效果不佳时,可通过选择与其遗传相关较高的次要性状来提高主要性状的预测可靠性;当个体缺失某个表型的记录时,可以参与到遗传评估中。本研究旨在评估最佳无偏线性预测(BLUP)与一步法基因组选择(ssGBLUP)两性状模型相比单性状模型的选择实施效果,为应对现实中个体可能缺失部分性状观察值提供参考依据。选择出生于2012—2019年的2 132头杜洛克猪为研究对象,利用DMU(v 6.0)软件中的单性状与多性状模型(BLUP,ssGBLUP)基因组选择方法探讨了验证群中不同比例的表型个体对生长性状(达100 kg体重时的日龄、背膘厚和眼肌面积)的预测可靠性的影响。结果:不同的验证群表型个体比例,单性状ssGBLUP对性状的预测可靠性均要优于单性状BLUP,两性状ssGBLUP模型的预测可靠性均要优于单性状模型;随着验证群中有表型的个体比例变化,两性状ssGBLUP与两性状BLUP模型对性状的预测可靠性均呈现提高的趋势。随着验证群中有表型的个体比例变化,两性状ssGBLUP模型对性状的预测可靠性明显高于两性状BLUP模型。研究表明两性状ssGBLUP的预测可靠性要优于两性状BLUP和单性状模型。在实际生产中,可以通过两性状ssGBLUP模型将个体纳入遗传评估中,为基因组选择在猪育种中的应用提供理论依据。
In genomic selection(GS), compared with the single-trait model, the multiple-trait model has many advantages. Firstly, it can utilize the information of genetic correlation between traits to improve prediction reliability. Secondly, when the direct selection of main traits fails, the prediction reliability can be improved by selecting secondary traits that are related to the main traits genetically. Thirdly, individuals lacking the phenotype records of some traits can still join the genetic evaluation. Our research was aimed to evaluate the selection performances of a two-trait model and a single-trait model, conducted by best linear unbiased prediction(BLUP) and single-step genomic best linear unbiased prediction(ssGBLUP), in order to provide a reference for dealing with the situation that some individuals that lacked the phenotype records of some traits in reality. 2 132 Duroc pigs born between 2012 to 2019 were selected as subjects of the present study. The impact of different proportions of phenotype individuals in the validation population on the prediction reliability of growth traits(age at 100 kg of growth, DAY;backfat thickness, BF;and loin eye area, LEA) were investigated using single and multiple traits(BLUP, ssGBLUP) methods in the DMU(v 6.0) software. The results were as follows: With different proportions of phenotype individuals in the validation population, the prediction reliability of ssGBLUP was better than that of BLUP. As the proportion of phenotype individuals in the validation population changed, it showed an increase tendency toward predictive reliability of the two-trait model conducted by BLUP and ssGBLUP, and the prediction reliability of the two-trait model conducted by ssG BLUP was significantly better than by BLUP. The prediction reliability of the two-trait model conducted by ssG BLUP was better than by BLUP and the single trait model. The present research suggested that,in breeding practice,individuals could join the genetic evaluation by using a two-trait ssG BLUP model,which provided theoretical reference for genetic selection in pig breeding.
作者
苏展勤
周隽
邵宝全
魏趁
邱定杰
张哲
张豪
李加琪
SU Zhanqin;ZHOU Jun;SHAO Baoquan;WEI Chen;QIU Dingjie;ZHANG Zhe;ZHANG Hao;LI Jiaqi(College of Animal Science,South China Agricultural University/National Engineering Research Center for Swine Breeding Industry,Guangzhou 510642,China;Fuqing Yongcheng Farming and Animal Husbandry Co.,Ltd.,Fuqing 350300,China)
出处
《畜牧与兽医》
CAS
北大核心
2023年第1期14-19,共6页
Animal Husbandry & Veterinary Medicine
基金
财政部和农业农村部:国家现代农业产业技术体系(CARS-35)。
