Simulating the impact of recombination rate on genomic selection breeding outcomes.

Recombination shuffles alleles during meiosis, driving genetic diversity and shaping the outcomes of breeding programs. By breaking the physical links between loci, recombination facilitates the creation of new allelic combinations that can be selected for to improve genetic gain. Increasing the recombination rate by methods such as genome editing has become a goal for accelerating breeding. However, the effect of increased recombination rate on a population scale on breeding programs is not fully understood. We therefore carried out simulations to determine the effect of recombination on genetic gain in a breeding program using phenotypic and genomic selection, respectively. We focused on how heritability, number of quantitative trait loci, recombination rate increase factor, marker density, and training frequency affect breeding success. We also tested whether it is possible to use historic training sets without changes in recombination rate and merge the pre and postrecombination populations to improve prediction accuracy and genetic gain in genomic selection. We found that increasing recombination is particularly beneficial for highly quantitative traits with low heritability. However, with genomic selection, increasing recombination requires a higher training frequency as well as an increased marker density to accelerate superiority over phenotypic selection in terms of genetic gain. Furthermore, our simulations show that maintenance of old training sets and merging of training sets with different recombination rates is possible, but a decrease in prediction accuracy is expected, favoring frequent training and high marker density under increased recombination rates.