Optimizing purebred selection to improve crossbred performance

Area of Science:

• Animal breeding and genetics
• Quantitative genetics
• Livestock production

Background:

• Crossbreeding enhances livestock production through heterosis and breed complementarity.
• Predicting crossbred performance (CP) often relies on purebred data due to scarcity of crossbred records.
• Accurate CP prediction requires accounting for non-additive genetic effects and environmental factors, influenced by the genetic correlation ($$) between populations.

Purpose of the Study:

• To investigate strategies for integrating purebred and crossbred data for optimal CP prediction.
• To evaluate single-step Genomic Best Linear Unbiased Prediction (ssGBLUP) and ssGBLUP with metafounders (ssGBLUP-MF) models under varying genetic correlation levels.
• To identify the most effective models for maximizing CP across different scenarios.

Main Methods:

• A two-way crossbreeding simulation was utilized.
• Scenarios explored genotyped individuals from purebred and crossbred populations.
• ssGBLUP and ssGBLUP-MF models were compared against BLUP with unknown parent group (BLUP-UPG).

Main Results:

• Prediction accuracy increased with higher $$ values across all scenarios.
• When incorporating genotypes from both parent breeds and crossbreds, ssGBLUP and ssGBLUP-MF showed similar accuracy, peaking at $$ < 0.5.
• At $$ = 0.8, ssGBLUP using only sire breed genotypes achieved the highest accuracy (73.2%).
• BLUP-UPG consistently showed lower accuracy than ssGBLUP and ssGBLUP-MF.
• Incorporating both crossbred and purebred genotypes improved prediction at lower $$ levels, while paternal genotypes were superior at high $$ levels.

Conclusions:

• The genetic correlation ($$) is a critical factor in predicting CP from purebred data.
• The optimal prediction model for CP is contingent on factors influencing $$.
• Further research is needed to develop models that optimize purebred selection for enhanced CP.