Estimation & Prediction

Evaluating the use of metafounders in an industry multibreed genetic evaluation in beef cattle

---

Abstract

The use of metafounders (MF) to model relationships in the base population has the potential of improving the accuracy of genetic evaluations in multibreed populations. This study aimed to evaluate the effectiveness of the MF method within an industry beef genetic evaluation program (INHERIT Selectâ„¢; https://beefgenetics.com/products/inherit-select). For this study, genotype information was available for 539,354 animals and 42,915 SNP after quality control. Pedigree information was available for 2,009,537 animals with 634,061 missing at least one parent. From those missing parents, 602,951 had breed composition information available either through genomic or pedigree. Three scenarios were investigated: (1) without the use of MF in the pedigree, and missing parents were set to zero; (2) 11 MF defined based on individual breeds; and (3) five MF defined based on breed origin: British, Continental, Dairy, Indicus, and Others. The predictive performance of the predictions from all scenarios were assessed by masking the phenotype of young animals for birth weight (BW, kg), fertility (FERT, number of calves produced by a cow up to the 8-calving, given that she had the first calf at 2 years of age) and carcass marbling (MARB, Beef Improvement Federation score). Adjusted phenotypes were obtained for all animals, incorporating all information available (phenotypes, pedigree, and genotypes). Cattle were ranked based on their Expected Progeny Differences (EPD) and divided into quartiles. The mean EPD, mean adjusted phenotype, and observed phenotype were computed for each quartile to assess the difference between the top and bottom-ranked cattle. The relationships from the gamma matrix among MF varied from 0.40 to 0.49 (MF11) and 0.40 to 0.47 (MF5). Results for MF11 were comparable or slightly better than those for MF5; therefore, only results for MF11 are presented. The correlations of EPD obtained from scenario 1 with those from scenario 2 ranged from 0.93 (BW) to 0.96 (MARB). The correlations of EPD from scenarios 1 and 2 with phenotypes were similar for MARB (0.81) and FERT (0.46) and slightly greater in scenario 1 for BW (0.67 vs. 0.65). Validation showed a small increase in differentiation between the top and bottom quantiles in scenario 1 compared to MF scenarios. Overall, the impact of MF on predictions depended on the trait and did not improve the prediction of phenotypes in relation to scenario 1 (i.e., not using MF) in the population investigated.

Keywords: 2026

How to Cite:

Gordo, D., Sanglard, L., Bernal Rubio, Y., Passafaro, T., Gonzalez-Peña, D., Ampofo, I., Fragomeni, B., Andersen, K. & Di Croce, F., (2026) “Evaluating the use of metafounders in an industry multibreed genetic evaluation in beef cattle”, World Congress on Genetics Applied to Livestock Production Digital Archive 2026(1): 2286408. doi: https://doi.org/10.31274/wcgalp.23989

Rights: 1