Genomic selection in admixed and crossbred populations.

Area of Science:

• Animal Genetics
• Quantitative Genetics
• Bioinformatics

Background:

• Genomic selection (GS) is crucial for livestock breeding, but typically studied in purebred populations.
• Real-world breeding data often involves crossbred or admixed populations with unknown breed composition.
• Using such data without accounting for breed composition can bias marker effect estimates due to stratification and admixture.

Purpose of the Study:

• To investigate the accuracy of genomic selection using training data from crossbred and admixed livestock populations.
• To evaluate the impact of marker density and breed composition on the reliability of genomic prediction equations.
• To determine if purebreds can be selected for crossbred performance using GS without pedigree or breed information.

Main Methods:

• Simulated a genome with varying marker densities (5-40 markers/cM).
• Created pure breeds and subsequently F(1), F(2), 3-, and 4-way crosses, and admixed populations.
• Used the Bayes-B method for marker effect estimation and calculated prediction accuracy (correlation of true with estimated breeding value) using a purebred validation set.

Main Results:

• Highest accuracy (0.79-0.85) was achieved when training data matched the purebred validation population.
• Crossbred and admixed training data yielded substantial accuracy (0.66-0.83), comparable to purebred training, especially with high-density markers.
• Accuracy decreased significantly if the target pure breed's genes were absent in the training population.
• Shorter haplotype segments in linkage disequilibrium were observed in crossbred/admixed populations.

Conclusions:

• High-density genomic selection is feasible and accurate using crossbred and admixed training data for all contributing pure breeds.
• Breed composition does not need explicit accounting for accurate GS prediction equations in mixed populations.
• Purebreds can be selected for crossbred performance using GS without pedigree or breed information, leveraging high-density markers.
• Shorter linkage disequilibrium segments in admixed populations offer potential for quantitative trait loci (QTL) fine mapping.