Bayesian prediction of breeding values for multivariate binary and continuous traits in simulated horse populations...

Area of Science:

• Animal Genetics
• Quantitative Genetics
• Statistical Genetics

Background:

• Accurate prediction of breeding values is crucial for genetic improvement in livestock.
• Mixed linear-threshold animal models are widely used for genetic evaluations.
• Integrating molecular genetic data can enhance prediction accuracy.

Purpose of the Study:

• To evaluate the properties of multivariate prediction of breeding values for categorical and continuous traits.
• To compare the efficacy of using phenotypic, molecular genetic, and pedigree information.
• To assess the impact of genetic marker scenarios on prediction accuracy.

Main Methods:

• Simulated data resembling Warmblood horse populations were used.
• Bayesian mixed linear-threshold animal models via Gibbs sampling were employed.
• Three scenarios of marker-quantitative trait locus (QTL) recombination rates and polymorphism information content (PIC) were simulated.

Main Results:

• Correlations between true and predicted breeding values ranged from 0.89-0.94 for continuous traits and 0.39-0.77 for binary traits.
• Combined phenotype and genotype data significantly outperformed phenotype data alone.
• Expected decrease in QTL trait prevalence was 3-12% with multi-trait selection and 9-17% with single-trait selection.

Conclusions:

• Phenotypic and highly informative genetic marker data should be integrated for breeding value prediction.
• This approach maximizes the reduction in prevalence of binary traits.
• The study provides a framework for genetic evaluation in complex trait scenarios.