Investigation on the metafounder concept in ssGBLUP based on a simulated cattle population



Single-step genomic best linear unbiased prediction (ssGBLUP) has become a popular tool for genetic evaluations in dairy cattle populations. The use of the metafounder (MF) concept allows better consideration of relationships within and between founder populations and ensures correct matching of pedigree and genomic relationships.

This study investigates the use of the MF concept in a simulated dairy cattle population where the base population consists of two related and inbred founder populations. The objectives are to compare genetic evaluations with and without MF and to investigate different methods of estimating MF parameters (Γ).

Results show that genetic evaluations using MF are less biased and less inflated compared to evaluations using unknown parent groups or not accounting for the different founder populations. However, testing different methods to estimate Γ revealed a tendency to overestimate the relationships within and between the founder populations, leading to an overestimation of pedigree relationships compared to the genomic relationships.

In summary, the MF concept in ssGBLUP is superior in this simulated scenario with two founder populations, but care must be taken when estimating Γ to ensure consistency between pedigree and genomic relationships. In general, these findings highlight the importance of considering relationships within and between founder populations in single-step genetic evaluations.