Genetic parameters via predictivity in large populations under strong genomic selection

Area of Science:

• Quantitative Genetics
• Animal Breeding
• Genomic Selection

Background:

• Accurate genetic parameter estimation requires comprehensive data, but current methods struggle with large genomic datasets and time-varying parameters.
• The development of efficient methods is crucial for advancing genomic selection and breeding strategies.

Purpose of the Study:

• To introduce and evaluate a novel method, Genetic Parameters via predictivity (GPP), for estimating genetic parameters.
• To assess GPP's performance across various data sizes, trait correlations, and time-varying scenarios.

Main Methods:

• GPP combines within- and across-trait predictivity formulas with a deterministic approach to estimate genomic breeding value accuracy.
• The method was tested on simulated datasets with positive and negative trait correlations under different population sizes (5,000 to 100,000 animals).

Main Results:

• GPP estimates closely matched realized values with minimal bias across simulated scenarios.
• The method demonstrated robustness to incorrect prior variance estimates and performed consistently across trait correlation types and data sizes.
• GPP exhibited a near-linear computational cost with increasing animal numbers, taking approximately 55 minutes for large datasets.

Conclusions:

• Genetic Parameters via predictivity (GPP) offers a fast, flexible, and accurate solution for estimating dynamic genetic parameters in large-scale genomic evaluations.
• GPP is suitable for datasets with complex genetic architectures, including positively or negatively correlated traits and varying genetic correlations over time.