Design of reference populations for genomic selection in crossbreeding programs

Area of Science:

• Animal Breeding and Genetics
• Genomic Selection
• Quantitative Genetics

Background:

• Genomic selection leverages crossbred (CB) data for purebred (PB) candidate selection.
• Reference populations typically include PB animals, but breeding goals often focus on CB performance.
• The study investigates the value of incorporating more CB individuals into reference populations.

Purpose of the Study:

• To evaluate the benefit of including varying proportions of CB animals in reference populations for genomic selection of PB candidates.
• To quantify the impact of CB inclusion on selection accuracy in crossbreeding programs.

Main Methods:

• A deterministic simulation study was conducted.
• The study used a pig breeding scheme with a moderately heritable trait.
• Reference population size was maintained at 6000 individuals.

Main Results:

• Selection accuracy for PB candidates increased from 0.49 to 0.60 when the reference population shifted from solely PB to a CB population twice the size of the PB line.
• The advantage of using CB over PB individuals increased linearly with the proportion of CB animals.
• Benefits diminished rapidly with higher genetic correlations (rPC) between PB and CB performance or when PB performance was part of the breeding objective.

Conclusions:

• Using CB individuals in reference populations offers significant advantages for genomic selection.
• These advantages are contingent on low genetic correlations between PB and CB performance.
• The inclusion of CB individuals is most beneficial when the breeding objective prioritizes CB performance over PB performance.