Genomic selection is used to predict breeding values of selection candidates using dense markers (Meuwissen et al. (2001)). In this approach one estimates an effect that each marker or haplotype has on the trait under study. Despite great success that genomic selection has had in animal breeding practice, the biological mechanisms underlying the trait remains largely unknown. Extension of the genomic models to include more biological information, such as gene pathways, can be interesting to increase the understanding of the selection mechanisms exploited, and to improve robustness, cross-population predictions, and also accuracy of genomic predictions. In this study we describe the inclusion of gene sets based on biological pathways in genomic models. We used gene sets that are associated to biological pathways from KEGG database (Kanehisa et al. (2008)) and analysed protein yield EBVs in Holstein bulls.

Keywords