Estimation & Prediction

Replacing Unknown Parent Groups with Metafounders in traditional pedigree-based U.S. Dairy Genetic Evaluations

---

Abstract

Metafounders (MF) extend unknown parent groups (UPG) by modelling relationships among base groups, offering a principled way to align pedigree and genomic bases in single-step genetic evaluations. We investigated the operational impact of redefining UPG as MF in the U.S. pedigree-based traditional BLUP evaluations run by Council on Dairy Cattle Breeding (CDCB), motivated by the need for smoother transitions when incorporating genotypes into single-step evaluation, after receiving and integrating MACE results from Interbull. We implemented an automated process to group unknown parents using the decision rules of A. Legarra and I. Aguilar (2024), which quantifies group informativeness and merge adjacent groups when weakly supported, improving stability and comparability of solutions. Then, we estimated a matrix Gamma of relationships among these groups (using base allele frequencies inferred during the routine imputation process at CDCB) and integrated them as MF in a pedigree-based traditional BLUP evaluations. Pedigree and phenotypic data from the official August 2024 multi-breed evaluation (Holstein, Jersey, Brown Swiss, Guernsey, Ayrshire, and Milking Shorthorn) were used, covering yield, somatic cell score, productive life, cow and heifer livability, cow health traits, and gestation length. We compared evaluation results from the MF approach with the official CDCB evaluations that model UPGs. Agreement between approaches was assessed via across-animal Pearson correlations of EBVs, changes in standard deviations, and cohort-wise genetic trends by birth year. Before comparison, EBVs were expressed on a common genetic base. Results showed the expected mean shifts for most traits under MF, consistent with the redefined base population, together with a general decrease in EBV standard deviation. For most trait groups, EBV correlations between MF-based and UPG-based evaluations were near unity for proven bulls and phenotyped cows with known parents. In most trait groups, genetic trends under MF differed only for older cohorts, altering slope and/or curvature and highlighting trait-dependent sensitivity to the modelling of base animals. Overall, replacing UPG with MF appears operationally feasible and scientifically coherent for U.S. pedigree-based traditional BLUP evaluations: it preserves rankings, and can stabilize interpretation when genotypes enter single-step evaluation. Our findings support adopting MF as the default base modelling strategy to improve continuity across traditional and genomic pipelines.

Keywords: 2026

How to Cite:

Savoia, S., Nicolazzi, E. & Legarra, A., (2026) “Replacing Unknown Parent Groups with Metafounders in traditional pedigree-based U.S. Dairy Genetic Evaluations”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2284555. doi: https://doi.org/10.31274/wcgalp.23589

Rights: 1