Bayesian divergence-based approach for genomic multitrait ordinal selection.

Area of Science:

• Plant breeding
• Quantitative genetics
• Bioinformatics

Background:

• Genomic selection for ordinal traits (e.g., disease resistance scores) presents challenges due to their discrete, ordered nature.
• Existing methods may not fully capture the complexities of multi-trait ordinal selection in plant breeding.

Purpose of the Study:

• To develop and evaluate a novel Bayesian divergence-based framework for multi-trait ordinal selection.
• To compare the performance of different decision-theoretic loss functions (KL divergence, Bhattacharyya distance, Hellinger distance) for genomic selection.

Main Methods:

• Implementation of a Bayesian divergence-based framework in the Multi-trait Parental Selection R package (MPS-R).
• Utilized Kullback-Leibler (KL) divergence, Bhattacharyya distance, and Hellinger distance to quantify distribution distances.
• Conducted extensive simulations across six scenarios with varying genetic correlations and heritability.
• Validated methods using real wheat disease resistance data.

Main Results:

• The Kullback-Leibler (KL) divergence loss function consistently achieved superior genetic gains, particularly under moderate heritability.
• The framework demonstrated robust performance across diverse genetic architectures and heritability levels.
• Validation with real wheat data confirmed the practical utility of the proposed methods.

Conclusions:

• The novel Bayesian divergence-based framework enhances selection efficiency for complex, multi-trait ordinal phenotypes in plant breeding.
• The MPS-R package provides a flexible and biologically grounded toolset for breeders.
• The Kullback-Leibler divergence is recommended for optimizing genetic gains in ordinal trait selection.