Evaluation Of Accuracy Of Genome Selection For Different Genetic Complexity Traits In Pigs

As an economic animal,porcine products enjoyed a great favor among the majority of consumers.Porcine breeding is an important field of animal husbandry in China.Traditional breeding methods take a long time and cost huge manpower and money resources,while genomic selection has the advantages of improving the accuracy of breeding value estimation,increasing annual genetic progress and reducing cost.Because of the complexity of the method of genomic selection and pig traits,it is necessary to compare the performance of genomic selection based on different hypothesis statistics methods according to different genetic complexity traits,choose the appropriate models,and formulate reasonable breeding strategies to accelerate the genetic improvement of pigs.In this study,data of teat number and backfat thickness at 100 kg of Yorkshire and teat number and age at 100 kg of Landrace were collected to estimate heritability of these traits firstly.And genome-wide association analysis was performed on the population.Finally,genomic selection was carried out for these traits with different genetic complexity.The results of this study are as follows:(1)Phenotypic data of 9412 Yorkshire pigs and 8215 Landrace pigs were collected and collated.The mean ± standard deviation of teat number of Yorkshire,backfat thickness at 100 kg of Yorkshire,teat number of Landrace and age at 100 kg of Landrace were 14.09±0.49,10.67±2.44,14.30±0.73 and 157.30±11.99,respectively.(2)Heritability of teat number of Yorkshire,backfat thickness at 100 kg of Yorkshire,teat number and age at 100 kg of Landrace by pedigree information was 0.150,0.312,0.210 and 0.409,respectively.Heritability of teat number of Yorkshire,backfat thickness at 100 kg of Yorkshire,teat number and age at 100 kg of Landrace by genome information was 0.076,0.237,0.125 and 0.126,respectively.By using comprehensive pedigree and genome information,the heritability of estimating teat number of Yorkshire,backfat thickness at 100 kg of Yorkshire,teat number and age at 100 kg of Landrace was 0.081,0.249,0.133 and 0.136,respectively.Heritability estimated by using pedigree information is higher than that by using comprehensive pedigree and genome information.Estimation of heritability by using comprehensive information is higher than that by genome information.(3)To analyze the traits with different genetic complexity of pigs,using genome-wide association analysis,the methods of GBLUP,ssGBLUP,BayesA,BayesB,BayesC,Bayesian LASSO,BSLMM and BayesR were carried out for comparing the accuracies and unbiasedness of genomic selection by a 5-fold and 10-fold cross-validation procedure on the base of the additive model,which combined with Porcine SNP50 K Beadchip.For teat number of Yorkshire,teat number and age at 100 kg of Landrace,ssGBLUP method was the best performance in genomic selection.BayesC method was the best model for backfat thickness at 100 kg of Yorkshire in genomic selection.ssGBLUP is more effective than GBLUP for pigs with different genetic complexity traits in genomic selection.(4)Trait with genetic complexity has an important influence on the effect of genomic selection.The prediction accuracy of the 8 methods showed a positive correlation with genetic ability estimation of traits.The model of genomic selection was studied by using actual population data to clear the performance of different selection methods of different genetic complexity traits and improve the stability of prediction,so as to provide reasonable reference for domestic pig genome selection.