Extension of the reduced animal model to single-step methods

Area of Science:

• Animal genetics
• Quantitative genetics
• Bioinformatics

Background:

• Traditional animal models (AMs) using best linear unbiased prediction (BLUP) face computational challenges with large equation systems.
• Reduced AM (RAM) simplified calculations by focusing on parents, followed by a back-solving step for progeny.
• Genomic information led to genomic BLUP (GBLUP) and single-step GBLUP (ssGBLUP) for integrated genetic evaluations.

Purpose of the Study:

• To extend the Reduced Animal Model (RAM) concept to single-step genomic best linear unbiased prediction (ssGBLUP) methods.
• To develop and evaluate novel RAM approaches for ssGBLUP that reduce computational burden.
• To assess the accuracy and efficiency of RAM-based ssGBLUP compared to full ssGBLUP.

Main Methods:

• Developed three distinct Reduced Animal Models (RAMs) for single-step GBLUP (ssGBLUP).
• Reduced the primary equation system from all animals to subsets including genotyped animals and specific nongenotyped groups.
• Employed a back-solving procedure to derive breeding values for remaining animals.

Main Results:

• All three developed RAMs for ssGBLUP produced results identical to the full ssGBLUP analysis.
• The RAM approaches significantly reduced the number of equations requiring direct solution.
• Demonstrated the feasibility of applying RAM principles to modern genomic evaluation models.

Conclusions:

• Reduced Animal Models (RAM) offer a computationally efficient alternative for single-step GBLUP (ssGBLUP).
• RAMs can mitigate the growing computational demands associated with large-scale genomic data in animal breeding.
• The study highlights the potential for RAM to be re-adopted for advanced genomic evaluations.