Genomic prediction using imputed whole-genome sequence in Australian Angus cattle

Abstract

<p>Using whole-genome sequence data in genomic prediction is expected to improve the predictive ability since the whole genome sequence may contain causal variants. This study aimed to compare the accuracy of genomic prediction with three densities of genotypes, 50k, high- density and wholegenome sequence. The genomic prediction was performed to estimate breeding values for selected growth and carcass traits in Australian Angus beef cattle. Genotype imputation was conducted to retrieve genotypes at high-density and whole-genome sequence level. The dataset was split into testing and reference group to compare the accuracy of breeding values obtained from different genotype densities and for animals with different degrees of relatedness to the reference. The prediction accuracies were similar across three different genotype densities for the traits studied. We found no substantial improvement in genomic prediction accuracy using the whole-genome sequence data in this study.</p>