Bayesian methods applied to genomic breeding value predictions for average daily weight gain in Nellore cattle

Abstract

The Nellore cattle is the most economically important breed reared in Brazil, because of their adaptation to the tropical environment, heat tolerance, parasite resistance and good reproductive performance, which are all factors that contribute to higher productivity. Most breeding programs have been based on body weight and weight gain traits, by using performance information for all related individuals in the population, giving greater genetic gain and increased inbreeding rates. In addition, genomic selection can reduce inbreeding rates, generation interval and maintenance of genetic variability, because of more accurate genomic estimated breeding values of young animals. Thus, appropriate methods can enable genomic selection and contribute to genetic improvement of Nellore cattle. Bayesian methods allow us to predict genomic breeding values by using different assumptions for marker effects and their variances. The aim of this study is to test different Bayesian methods to estimate marker effects and equations for predicting GEBVs. (AU)