| Genomic selection is an important genetic improvement method,which is of great significance for genetic evaluation and has been widely used in different species.The size of reference population is an important factor affecting the whole genome genetic evaluation in GS.In the multi-population genetic evaluation,GS is usually carried out in populations with similar genetic background,while in actual production,the population with similar genetic background is relatively small.Therefore,it is necessary to explore the feasibility and strategy improving the accuracy of Genomic estimated breeding value by combining population and expanding the size of population.Teats number is an important reproductive trait,which belongs to medium heritability trait.Therefore,this study intends to expand the scale of reference population improving the accuracy of total teats number GEBV estimation and analyzes the optimal strategy by combining the Danish Yorkshire and American Yorkshire.Firstly,determining the genetic correlation of total teats number between Danish Yorkshire and American Yorkshire by analyzing the genetic parameters of the two populations.Secondly,the mixed linear model and Bayesian C model were combined to determine whether there are difference in genetic background.Finally,considering the size of reference population size and genetic background factors,designing four genetic evaluation models to predict GEBV:(1)it is assumed that the genetic background is different,and the size of the reference population is not expanded,denoted as model M1;(2)assuming no difference in genetic background,expanding the scale of the reference population,denoted as model M2;(3)adding the line effect to model M2,denoted as model M3;(4)assuming there are difference in genetic background,enlarging the reference population,denoted as M4.The main results are as follows: 1.Results of genetic parameters estimation for total teats number in Danish Yorkshire and American Yorkshire(1)Estimation of heritability in Danish Yorkshire with 10060 phenotype information and 1497 genotype information indicated that: the heritability was 0.167±0.02 in traditional BLUP and 0.095±0.03 in GBLUP;(2)Estimation of heritability in American Yorkshire with 13060 phenotype information and 1793 genotype information indicated that: the heritability was 0.423±0.02 in traditional BLUP and 0.402±0.04 in GBLUP;(3)Genetic correlation of total teats number between Danish Yorkshire and American Yorkshire was 0.789±0.328.2.Results of GWAS in Danish Yorkshire and American Yorkshire(1)By using LMM method,we found that the significant SNP associating with total teats number were all located on chromosome 7 in Danish Yorkshire and American Yorkshire,but the regions were different.The area was 97.19~97.80 Mb in Danish Yorkshire and 97.08~98.09 Mb in American Yorkshire;By using Bayes C method,we found that associated with total teats number in Danish Yorkshire was located one chromosome 7,while,in American Yorkshire was located on chromosome 3,4,7,10 and 11.The above results showed that the chromosome region controlling the total teats number between Danish Yorkshire and American Yorkshire were not completely the same,and the genetic background was not completely the same.(2)Combining analysis of above genome wide association analysis results and then by linkage disequilibrium analysis,we found 9 candidate genes in Danish Yorkshire,including PTGR2,FAM161 B,ENTPD5,LIN52,ABCD4,VRTN,SYNDIG1 L,NPC2 and LTBP2 genes;we found 13 candidate genes in Danish Yorkshire,including ELMSAN1、PTGR2、FAM161B、ENTPD5、LIN52、ABCD4、VRTN、NPC2、LTBP2、FCF1、YLPM1 、PROX2、DLST genes.3.Results of GEBV prediction in total teats number trait using four models(1)Results of prediction in Danish Yorkshire indicated: The accuracy and unbiasedness of GEBV by model M4 were the best,followed by model M1,and the accuracy and unbiasedness of GEBV predicted by model M2 and M3 were the worst.(2)Results of prediction in Danish Yorkshire indicated: Among the four models,the accuracy and unbiasedness of GEBV predicted by model M4 were the best.In the accuracy of GEBV,model M2 and M3 were only next to model M4,and model M1 had the worst performance.In unbiasedness of breeding value,model M1 was only next to model M4,while model M2 and M3 had the worst performance.Conclusions:(1)The candidate genes controlling total number of teats were not completely the same in Danish and American Yorkshire and there were the same and unique association region,indicating that the genetic background in two line were different;(2)The accuracy of GEBV improved by combining Danish and American Yorkshire;(3)The model M4 considering the size of reference population and genetic background,which was the best strategy for genome genetic evaluation. |
PDF Full Text Request