Interpretive Summary: Genomic predictions impacts when new information is added in a population of American Angus cattle
By: Anne Kamiya, MS
Predictive models are extremely important for maximizing efficiency and productivity in cattle breeding. The accuracy and stability of genomic predictive models are dependent on population parameters and how much phenotypic and genotypic data is included. In general, the more data available when creating a model, the higher the stability of that model will be.
In this recent Journal of Animal Science study, researchers looked at the estimated breeding values (EBV) and genomic EBV (GEBV) in a population of American Angus cattle for one year. The dataset used for this study was extremely large. Their primary goal was to determine how stable the models were when new genotypic and phenotypic information were added at monthly intervals. They also wanted to evaluate how new genotypic data impacted the stability of GEBV.
Best linear unbiased prediction (BLUP) model was used to evaluate EBV and single-step genomic best linear unbiased prediction model (ssBLUP) with the algorithm for proven and young (APY) was used to evaluate GEBV. Average absolute changes in GEBV were larger than EBV, but extreme maximum absolute changes were higher in EBV. Overall, the authors reported that genomic data yielded accurate estimates despite addition of new information. The results of this study suggest both EBV and GEBV remain accurate and stable when new genomic and phenotypic data are added to an existing large, genotyped population. More studies on the impacts new information may have on genomic predictions in different cattle populations may be warranted.
The full article, Changes in genomic predictions when new information is added, is available on the Journal of Animal Science website.