Single-step genomic analysis of dominance in a crossbred pig population

When all animals are genotyped, developments focused on dominance are well derived. However, this is rare in practice and adaptations are required. The dominant pedigree-based and genomic matrices should be compatible in scale to be combined. We aimed to show how gene content flows from genotyped to non-genotyped animals. Scenarios where 0% (D), 20% (C20), 40% (C40) 60% (C60), 80% (C80), and 100% (G) of animals were genotyped were assumed in a crossbred pig population. Dominance variances were more sensitive by choosing G rather than D. D and C60 presented the highest dominance variances whereas G showed the lowest. The prediction accuracies increased as the number of genotyped animals increased. C40 provided less inflated dominance breeding values (1.01), and 55% of accuracy. Dominance relationships flew from genotyped to non-genotyped animals but single-step extension might worked because we used an approximation of D where inbreeding is ignored and due to simple family structure in our dataset. Thus, our approach might be useful but is not general as the original derivations of single-step. Genotyping strategies can be optimized based on the proportion of genotyped animals where 40% seemed to be optimal.