Implications of Genetic Architecture on the Efficacy of Genomic Selection

A simulation model is described which has been constructed to address the issue of how true underlying genomic architecture might impact on the efficacy of genomic selection. A current specific focus of the model is on how epistatic genetic architectures might impact on the added value expected from increasing the density of SNP markers. Results to date suggest that genomic selection has greater superiority over BLUP genetic prediction under the additive genetic architecture simulated relative to an epistatic architecture with similar heritability. While we expect marker density to improve accuracy under GBLUP with some additive genetic architectures, our simulation results suggest that this may not happen with comparable (in terms of narrow sense trait heritability) genetic architectures with epistatic gene action contributing to both additive and non-additive genetic variance.