Simulation Study on the Impact of Accuracy of Estimated Genetic Relationship Matrices on Predicting Genotype Performance

• Activity Number: 32 - Longitudinal/Correlated Data I
• Type: Contributed
• Date/Time: Sunday, July 29, 2018 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): MINGZHU SUN* and Vivi Arief and Ian DeLacy and Kaye Basford and Wen-Hsi Yang
• Companies: The University of Queensland and UNIVERSITY OF QUEENSLAND and UNIVERSITY OF QUEENSLAND and UNIVERSITY OF QUEENSLAND and UNIVERSITY OF QUEENSLAND
• Keywords: mixed model; genetic relationship matrix; plant breeding; field trials

Abstract

Using pedigree or molecular marker based data in plant breeding has become more prevalent. These data contain genetic dependency among genotypes in plant breeding trials and have been used to estimate genetic relationship matrices for prediction of genotype performance. Although there are many estimators for this dependency, the estimated matrices being used might not capture the true relationship among genotypes due to limited marker coverage, incomplete pedigree information, or the recording error. There is concern that using inappropriate genetic relationship matrices might result in misleading prediction of genotype performance. However, the influence of using such inappropriate matrices for prediction has not been studied. In this study, we use simulation to evaluate the influence of using various relationship matrices with various levels of accuracy.