gEBVs

GEBV stands for Genomic Estimated Breeding Value, a prediction of an individual’s additive genetic merit for a trait based on genomic information. Unlike traditional EBV, GEBV uses DNA marker data to capture effects from many loci across the genome.

In genomic selection, dense SNP data are combined with phenotypes and pedigree to estimate small effects of numerous variants. A training population with known genotypes and phenotypes calibrates a predictive model, which is then applied to selection candidates.

Common approaches include Genomic BLUP (GBLUP), which uses a genomic relationship matrix, and Bayesian methods (BayesA/B/Cπ, BayesR) that assign priors to SNP effects. The GEBV for an individual is the sum of its estimated marker effects.

GEBVs enable earlier, more accurate selection and can increase genetic gain by reducing generation intervals. They are widely used in dairy and beef cattle, swine, poultry, and increasingly in crops.

Limitations include the need for large, relevant training populations; accuracy declines for distant populations or low-heritability traits. Transferability across breeds or populations is limited, and predictions carry uncertainty that should be considered in selection decisions.

GEBVs complement traditional EBVs by incorporating genomic information while remaining interpretable as additive merit estimates. They are typically expressed in the same units as EBV, facilitating integration into breeding programs.