Efficient implementation of multitrait random regression test-day models with external information for dairy cattle...

Area of Science:

• Animal Breeding and Genetics
• Quantitative Genetics
• Dairy Cattle Genomics

Background:

• Random regression models (RRM) and single-step genomic best linear unbiased prediction (ssGBLUP) are standard for dairy cattle genomic evaluations.
• Efficient implementation of RRM with ssGBLUP is crucial for national genetic evaluations.
• Integrating international data enhances the accuracy and scope of genomic predictions.

Purpose of the Study:

• To efficiently implement RRM combined with ssGBLUP for national dairy cattle genetic evaluations.
• To integrate external multicountry evaluation approach (MACE) breeding values into the national system.
• To assess the performance and accuracy of the implemented genomic evaluation system.

Main Methods:

• Utilized a large dataset of 30 million test-day records and 2.5 million pedigree animals from the Czech Holstein population.
• Employed reduced genetic groups and the algorithm for proven and young (APY) to enhance model convergence and computational speed.
• Solved mixed model equations using preconditioned conjugate gradient and incorporated external MACE deregressed proofs (DRP) weighted by effective record contributions (ERC).

Main Results:

• The implemented ssGBLUP with RRM achieved convergence and demonstrated desirable validation statistics (bias near zero, high dispersion, strong correlations).
• The algorithm for proven and young (APY) accelerated the ssGBLUP process by 10-fold.
• Integration of MACE information improved the correlation between national and international reliabilities for both BLUP and ssGBLUP.

Conclusions:

• The application of ssGBLUP to a multitrait RRM is feasible for national dairy cattle evaluations.
• The system efficiently integrates external MACE information, leading to highly accurate genomic estimated breeding values (GEBV).
• The developed approach provides a robust and computationally efficient genomic evaluation system for dairy populations.