Abstract

Single-trait and multi-trait GBLUP models were compared with a recently developed Bayesian model, BayesAS, for their accuracy of genomic prediction. For this purpose, a simulation study were performed considering two correlated traits with heritability of 0.4 and 0.1, various numbers of records for the trait with heritability of 0.1, and three scenarios with regard to varying local and genome-wide (co)variances between the traits. In BayesAS analysis, varying SNP bin sizes defined as a genomic region were used. BayesAS models performed better than GBLUP models across the three scenarios for the high heritability trait in both single-trait and multi-trait analysis. For the trait with lower heritability, the advantage of BayesAS models depended on the number of animals with phenotypic records and the SNP bin size. When increasing the SNP bin size up to 200, multi-trait BayesAS model was able to capture the heterogeneous (co)variance and thus allowed better use of information from correlated trait and increased accuracy of prediction for the trait with low heritability and/or small number of records. Key words : genetic correlation, heterogeneous (co)variances, multi-trait genomic prediction, BayesAS model ––––––––––––––––––––––––––-Page 1––––––––––––––––––––––––––- ? Thank you for using PDF Editor 6 Professional. You can only convert up to 5 pages in the trial version. To get the full version, please purchase the program here: http://cbs.iSkysoft.com/go.php?pid=2982&m=db

Mahlet Teka Anche, Guosheng Su, Emre Karaman, Luc L Janss, Mogens Lund

Proceedings of the World Congress on Genetics Applied to Livestock Production, Volume Methods and Tools - Models and Computing Strategies 2, , 284, 2018
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.