Genomic Selection Of Threshold Traits In Beef Simmental Cattle

Genomic selection(GS)is an important method of animal and plant genetic breeding.By estimating the molecular marker effect covering the whole genome,thegenomic estimated breeding value(GEBV)of individuals can be obtained,and finally the candidate individuals can be evaluated.At the same time,compared with the traditional breeding technology,the accuracy of breeding value estimation is higher for the characters of limited,low heritability and difficult to measure.Threshold traits such as meat color,fat color and marbling score are important economic traits that affect beef cattle breeding efficiency.Accurate genetic evaluation of these traits is the basis of continuously improving population breeding efficiency.Evaluation model is one of the most important factors that affect the accuracy of genome selection.Most of the current genetic evaluation models assume that the traits are continuous and consistent with normal distribution.However,for the threshold traits of discrete distribution,the premise assumption of conventional linear model can not be established.In order to improve the accuracy of breeding value estimation for threshold traits,linear model and threshold trait model were used to estimate genomic parameters and genomic breeding value for simulation data and Simmental cattle discrete traits breeding data.Then,the whole genome association analysis of color,fat color and marbling score of Chinese Simmental beef population was carried out using threshold trait model.Through the above analysis,we found that:(1)in all the simulation data sets,BayesTHB has the highest accuracy and BayesA has the lowest accuracy.The accuracy of BayesTH model is higher than Bayes model in estimating the breeding value of threshold character genome.The accuracy of each model decreased with the increase of QTL number,and increased with the increase of heritability,number of trait categories and incidence.(2)When using GBLUP and GLMM models to estimate the genetic parameters of Simmental beef color,fat color and marbling score respectively,it was found that for the above three threshold traits,the estimated values of genetic variance and heritability in GLMM model were higher than those in GBLUP model,and the standard deviation of estimated values was smaller.(3)In the real data,seven evaluation models were used to estimate the genetic breeding value of three threshold traits: meat color,fat color and marbling score.The results showed that the accuracy of BayeTH model was higher than Bayes and GBLUP model.(4)In genome-wide association analysis,seven candidate gene regions were screened by BayesTHB model.Through further localization,TMEM236 can affect both meat color and fat color.Based on the above results,the conclusions are as follows:(1)BayesTHB has the highest accuracy in the simulation data,which is more suitable for the estimation of breeding value of threshold traits.(2)Compared with GBLUP model,GLMM model can segment additive genetic variance more accurately.(3)In the real data,BayesTHB has the highest accuracy in genetic evaluation of meat color,fat color and marbling score,which is more suitable for breeding value estimation of threshold traits.(4)Based on the evaluation results of the model and real traits in this study,it is concluded that when the number of threshold characters is greater than or equal to 7,it can be treated as a continuous trait approximately.(5)TMEM236 gene affects both meat color and fat color,which may have multiple validity.