پردیس کشاورزی و منابع طبیعی دانشگاه تهران Iranian Journal of animal Science 2008-4773 48 4 2018 02 20 Evaluation of the effectiveness of bootstrap aggregating sampling technique in the accuracy of genomic best linear unbiased prediction method Evaluation of the effectiveness of bootstrap aggregating sampling technique in the accuracy of genomic best linear unbiased prediction method 573 584 65830 10.22059/ijas.2018.248547.653596 FA Khabat Kheirabadi Ph.D. Candidate, Department of Animal Science, Khuzestan Agricultural Sciences and Natural Resources University, Iran Jamal Fayazi Associate Professor, Department of Animal Science, Khuzestan Agricultural Sciences and Natural Resources University, Iran Hedayatollah Roshanfekr Professor, Department of Animal Science, Khuzestan Agricultural Sciences and Natural Resources University, Iran Rostam Abdollahi Arpanahi Assistant Professor, Department of Animal and Poultry Science, Aburaihan Campus, University of Tehran, Iran Journal Article 2018 02 05 In order to increase the accuracy of genomic best linear unbiased prediction method (GBLUP), bootstrap aggregating sampling (bagging) technique was applied. In this order a genome consisted of 10,000 bi-allelic single nucleotide polymorphism (SNP) over ten chromosomes, with 100 cM length each, was simulated. To generate linkage disequilibrium (LD) between SNPs and quantitative trait loci (QTL), random mating was simulated for 100 generations between 100 individuals (50 males and 50 females). Then in generation 101 (reference population) number of individuals increased to 1000 or 2000 and their phenotypes were also simulated. Then the marker effects were estimated in this population using GBLUP method or combined this method with bagging technique (BGBLUP). By using these regression coefficients and according to the genotype markers for juvenile individuals in generations 102 to 105, called validation population which had no phenotype, genomic breeding values were predicted. According to the finding of this research, the accuracies of genomic breeding values of GBLUP method were higher than those for BGBLUP (<em>p</em> > 0.05) and about the first testing set (102 generation) and regardless of QTL effects with a population of 1000 (or 2000) observations in the reference set, the range of GBLUP accuracy was 0.339±0.049 (0.412±0.042) for a trait with 0.05 heritability to 0.728±0.015 (0.783±0.015) for a trait with 0.65 heritability, whereas the accuracy of BGBLUP method were varied between 0.338±0.047 (0.411±0.042) to 0.725±0.016 (0.780±0.015). In order to increase the accuracy of genomic best linear unbiased prediction method (GBLUP), bootstrap aggregating sampling (bagging) technique was applied. In this order a genome consisted of 10,000 bi-allelic single nucleotide polymorphism (SNP) over ten chromosomes, with 100 cM length each, was simulated. To generate linkage disequilibrium (LD) between SNPs and quantitative trait loci (QTL), random mating was simulated for 100 generations between 100 individuals (50 males and 50 females). Then in generation 101 (reference population) number of individuals increased to 1000 or 2000 and their phenotypes were also simulated. Then the marker effects were estimated in this population using GBLUP method or combined this method with bagging technique (BGBLUP). By using these regression coefficients and according to the genotype markers for juvenile individuals in generations 102 to 105, called validation population which had no phenotype, genomic breeding values were predicted. According to the finding of this research, the accuracies of genomic breeding values of GBLUP method were higher than those for BGBLUP (<em>p</em> > 0.05) and about the first testing set (102 generation) and regardless of QTL effects with a population of 1000 (or 2000) observations in the reference set, the range of GBLUP accuracy was 0.339±0.049 (0.412±0.042) for a trait with 0.05 heritability to 0.728±0.015 (0.783±0.015) for a trait with 0.65 heritability, whereas the accuracy of BGBLUP method were varied between 0.338±0.047 (0.411±0.042) to 0.725±0.016 (0.780±0.015). https://ijas.ut.ac.ir/article_65830_a2bc5c798dfb59b5d25d7e5bc413b4f6.pdf