Accuracy of Genomic Prediction for Merino Wool Traits Using High-Density Marker Genotypes

Abstract

High-density (HD) marker genotypes could increase the accuracy of genomic prediction by providing stronger linkage disequilibrium (LD) between markers and quantitative trait loci affecting a trait, especially in populations with a high genetic diversity such as Australian Merino sheep. The aim of this study was to compare the accuracy of genomic prediction for Merino yearling and adult wool traits based on observed and imputed 600K single nucleotide polymorphism (SNP) marker genotypes with the accuracy based on moderate-density (50K) marker genotypes. Genomic best linear unbiased prediction (GBLUP) and a Bayesian approach (BayesR) were used as prediction methods. Results showed a small relative increase in accuracy between 2 to 15% (of the previous accuracy) when using a HD marker set. The results of BayesR were on average similar to GBLUP. Considerably higher (up to 25% relative increase) in prediction accuracy was observed for animals with lower genomic relationship to the reference population.