A sampling method for estimating the accuracy of predicted breeding values in genetic evaluation

Area of Science:

• Animal breeding and genetics
• Quantitative genetics
• Statistical genetics

Background:

• Accurate estimation of breeding values is crucial for genetic improvement in livestock.
• Traditional methods may not fully capture the uncertainty associated with estimated breeding values (EBVs).
• The animal model is a standard in modern genetic evaluations.

Purpose of the Study:

• To present a novel sampling-based method for estimating the accuracy of EBVs.
• To assess the performance and applicability of this method across different genetic models.
• To explore the method's utility in genetic connectedness studies.

Main Methods:

• A simulation-based approach was used to estimate empirical variances of true and EBVs.
• The method was validated using a small dataset from the Parthenaise cattle breed.
• Application to a large French Salers dataset for muscle development score evaluation (IBOVAL).

Main Results:

• The coefficient of determination estimated by the method converged to true values in validation.
• The method demonstrated applicability to various genetic models, including sire, multivariate, and maternal effects.
• Off-diagonal coefficients of the inverse of mixed model equations were successfully supplied.

Conclusions:

• The sampling-based method provides a robust way to estimate EBV accuracy.
• While computationally intensive for large datasets, it offers flexibility and insights into genetic connectedness.
• This method enhances the reliability of genetic evaluations and aids in strategic breeding program design.