USWBSI Abstract Viewer

2023 National Fusarium Head Blight Forum


Variety Development and Host Resistance (VDHR)

Poster # 508

Investigating the Potential of Weighted Genomic Relationship Matrix in Optimizing Prediction Accuracy of Deoxynivalenol Accumulation in Barley

Authors & Affiliations:

Adenike D. Ige 1 and Kevin P. Smith 1
1. University of Minnesota, Department of Agronomy and Plant Genetics, St. Paul, Minnesota
Corresponding author: Kevin P. Smith, smith276@umn.edu

Corresponding Author:

Adenike Ige
aige@umn.edu

Abstract:

Fusarium head blight (FHB) is a devastating fungal disease of barley that negatively affects yield and grain quality. Barley grains infected with Fusarium graminearum accumulate the mycotoxin; Deoxynivalenol (DON), and if concentrations exceed the detectable levels, it can cause food poisoning and drastically reduce global market value. Growing resistant genotypes is the most environmentally friendly and effective strategy to manage the disease. Genomic selection is an important genome-based breeding approach that can identify resistant genotypes by predicting their genetic merits with high accuracy. It can save costs associated with DON quantification and the establishment of specialized irrigated nurseries for FHB disease screening. Despite these potential advantages, it is important to optimize the genomic prediction models to accurately predict DON accumulation. GBLUP is one of the preferred prediction models for routine genomic evaluations of genotypes because it exploits information from relatives for predictions and has low computational demand. However, its assumption of the equal contribution of all loci to the genetic variance is violated for traits such as DON, which is controlled by a defined number of quantitative trat loci. Weighted GBLUP (WGBLUP), which incorporates a trait-specific relationship matrix in place of the conventional G-matrix, is a modified approach for predicting breeding values and can account for deviation from the GBLUP assumption. Therefore, the objectives of this study were to (i) identify the optimal weighting method for WGBLUP using results from genomic prediction models (RRBLUP and BayesC), and (ii) compare the predictive ability of WGBLUP models and other models in the training and validation populations of barley. Existing phenotype and genotype data from the University of Minnesota barley breeding populations were used. This study provides useful information on the performance of WGBLUP and other models, as well as the appropriate weighting strategy for the optimal performance of WGBLUP.


©Copyright 2023 by individual authors. All rights reserved. No part of this abstract or paper publication may be reproduced without prior permission from the applicable author(s).