Multi-trait genomic selection for combined leaf and stem rust resistance in wheat breeding
- 1Czech Agrifood Research Centre, Drnovská 507, Prague 6, Ruzyně, 161 00, Czech Republic; Corresponding author. Czech Agrifood Research Centre, Drnovská 507, Prague 6, 161 00, Ruzyně, Czech Republic.
- 2Czech Agrifood Research Centre, Drnovská 507, Prague 6, Ruzyně, 161 00, Czech Republic
- 3Selgen, Jankovcova 24/18, Prague 7, Czech Republic
Abstract
Genomic selection (GS) is a promising approach for improving wheat resistance to rust diseases. This study focused on two major rust diseases: leaf rust caused by Puccinia triticina and stem rust caused by P. graminis f.sp. tritici. The training population consisted of 186 wheat accessions with a wide range of resistance levels, while the breeding population consisted of 103 wheat breeding lines. Both populations were genotyped using the Diversity Arrays Technology (DArTseq) platform. The training population was phenotyped under field conditions at two locations over at least three years. In addition to visual assessment of disease symptoms, pathogen content in host plants was quantified using real-time PCR. Six genomic prediction models were tested in a single-trait framework, and the Bayes A model showed the best performance according to the selected decision criteria. Prediction accuracies ranged from 0.79 to 0.75 for stem rust and from 0.42 to 0.38 for leaf rust. Bayesian Ridge Regression (BRR) was selected for multi-trait GS. A very high correlation (0.98–1.00) was observed between genomic estimated breeding values obtained from the single- and multi-trait analyses. Twenty-one promising breeding lines were identified using a Kullback–Leibler loss function combined with average genetic similarity. The selected lines showed a potential improvement in resistance of 16% for stem rust and 5% for leaf rust. GS, especially in multi-trait arrangement, proved to be a good tool for accelerating breeding efficiency for combined and complex traits.
Related
- Integrating Abiotic Stress Adaptation and Disease Resistance in Climate-Resilient Crops: Pangenomics, Genomic Selection, and AI-Driven Breeding
- Multi-trait selection of tropical maize hybrids for combined productivity, agronomic stability, and aflatoxin resistance
- Combining different selection criteria into breeding strategies for short-term recurrent genomic selection
- A unified decision framework combines phenotypic and genomic information for nonlinear selection in plant breeding
- Mutation breeding for quantitative traits in the era of genomic selection
BibTeX
@article{leisovasvobodova2026multitrait,
title = {{Multi-trait genomic selection for combined leaf and stem rust resistance in wheat breeding}},
author = {Leona Leišová-Svobodová and Alena Hanzalová and Ondřej Veškrna and Michaela Jungová and Tereza Sovová and Pavel Svoboda},
journal = {Journal of Agriculture and Food Research},
year = {2026},
month = oct,
doi = {10.1016/j.jafr.2026.103381},
url = {https://doi.org/10.1016/j.jafr.2026.103381},
note = {CC BY},
}
RIS
TY - JOUR
TI - Multi-trait genomic selection for combined leaf and stem rust resistance in wheat breeding
AU - Leona Leišová-Svobodová
AU - Alena Hanzalová
AU - Ondřej Veškrna
AU - Michaela Jungová
AU - Tereza Sovová
AU - Pavel Svoboda
PY - 2026
DA - 2026/10/30
JO - Journal of Agriculture and Food Research
DO - 10.1016/j.jafr.2026.103381
UR - https://doi.org/10.1016/j.jafr.2026.103381
AB - Genomic selection (GS) is a promising approach for improving wheat resistance to rust diseases. This study focused on two major rust diseases: leaf rust caused by Puccinia triticina and stem rust caused by P. graminis f.sp. tritici. The training population consisted of 186 wheat accessions with a wide range of resistance levels, while the breeding population consisted of 103 wheat breeding lines. Both populations were genotyped using the Diversity Arrays Technology (DArTseq) platform. The training population was phenotyped under field conditions at two locations over at least three years. In addition to visual assessment of disease symptoms, pathogen content in host plants was quantified using real-time PCR. Six genomic prediction models were tested in a single-trait framework, and the Bayes A model showed the best performance according to the selected decision criteria. Prediction accuracies ranged from 0.79 to 0.75 for stem rust and from 0.42 to 0.38 for leaf rust. Bayesian Ridge Regression (BRR) was selected for multi-trait GS. A very high correlation (0.98–1.00) was observed between genomic estimated breeding values obtained from the single- and multi-trait analyses. Twenty-one promising breeding lines were identified using a Kullback–Leibler loss function combined with average genetic similarity. The selected lines showed a potential improvement in resistance of 16% for stem rust and 5% for leaf rust. GS, especially in multi-trait arrangement, proved to be a good tool for accelerating breeding efficiency for combined and complex traits.
SN - 2666-1543
ER -