Imputation accuracy for missing alleles in crossbred beef cattle

Abstract

<p>Establishing a consistent set of markers through the imputation of various marker densities is crucial in building the genomic relationship matrix for genomic prediction. However, imputation in crossbred populations presents challenges. This paper investigated the imputation accuracy of purebred and crossbred Brangus, Simmental and Wagyu cattle. The imputation was carried out independently within each breed using a population-based approach. A reference population of 3,000 randomly selected purebred and crossbred animals with medium-density markers was used in each population to impute the target population with low-density markers (10,000 markers), and this process was repeated five times. On average, imputation accuracies higher than 0.9 were estimated for all three populations. However, the accuracy decreased as the relationship to the reference population within each breed decreased.</p>