Technical note: Equivalent genomic models with a residual polygenic effect.

Area of Science:

• Animal breeding and genetics
• Quantitative genetics
• Genomic evaluation methodologies

Background:

• Routine genomic evaluations commonly use Genomic Best Linear Unbiased Prediction (GBLUP) or Single Nucleotide Polymorphism (SNP) BLUP models.
• These models predict genetic merit but face challenges due to incomplete linkage disequilibrium between SNP markers and causal variants.
• A residual polygenic effect is often included to capture unexplained genetic variance.

Purpose of the Study:

• To prove the equivalence of multi-step GBLUP and SNP BLUP models, including residual polygenic effects.
• To verify this equivalence for both phenotyped reference animals and young genotyped animals without phenotypes.
• To extend the proof of equivalence to single-step genomic evaluation and SNP effect estimation.

Main Methods:

• Mathematical proofs were derived to demonstrate model equivalence.
• The study focused on comparing GBLUP and SNP BLUP under multi-step and single-step evaluation frameworks.
• Formulas for converting GBLUP estimated breeding values into direct genomic values and residual polygenic effects were developed.

Main Results:

• Equivalence between multi-step GBLUP and SNP BLUP models was proven for reference animals with residual polygenic effects.
• This equivalence was confirmed for young genotyped animals lacking phenotypes.
• The study demonstrated that this equivalence extends to single-step genomic evaluations, yielding identical predictions for all genotyped animals and for SNP effect estimation.

Conclusions:

• Both multi-step and single-step GBLUP and SNP BLUP models, when incorporating residual polygenic effects, provide equivalent genomic predictions.
• The findings simplify genomic evaluation strategies by confirming the interchangeability of these models.
• This equivalence enhances the reliability of genomic predictions for diverse animal populations and breeding scenarios.