Theoretically, the more data in the training set, the better the accuracy of the predicted breeding values. However, adding distant generations to the training data will introduce computational burden, with perhaps limited contributions to prediction. The objectives of this study were to compare the accuracy of marker-based and pedigree-based models and to evaluate the optimum number of training generations required to most accurately predict EBV in a commercial layer breeding line. On average, accuracies of EBV based on markers were higher than accuracies based on pedigree. Accuracies of all methods initially increased with successive increases in the number of generations of training data, but slightly dropped or reached an asymptote when including training generations far apart from validation. The divergence in gene frequencies in each generation, genotype by environment interactions, and selection over generations might be the causes of these decreases in accuracy.

Zi-Qing Weng, Anna Wolc, Rohan L Fernando, Jack CM Dekkers, Jesus Arango, Janet E Fulton, Petek Settar, Neil P O'Sullivan, Dorian J Garrick

Proceedings of the World Congress on Genetics Applied to Livestock Production, Volume Species Breeding: Poultry, , 326, 2014
Download Full PDF BibTEX Citation Endnote Citation Search the Proceedings

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.