Interpretive Summary: Hidalgo Accuracy Genomic Prediction Broilers
By: Caitlin Vonderohe
Accuracy of genomic predictions is one of the most important facets of an animal breeding program because it has a direct relationship with the response of the animal population (in growth, performance, reproductive capacity) to genetic selection. This depends on how much genetic variance is controlled by single nucleotide polymorphisms (SNPs), which is, in turn, dependent on the data used to estimate the SNPs, the heritability of the traits being assessed, and the statistical methods used when making predictions.
In poultry, the accuracy genomic prediction has historically remained relatively stable when animal with known genotypes and phenotypes were added to the breeding population, but was dramatically reduced when no new animals were included, and as a result the response to selection was smaller than expected. A recent paper by Hidalgo et al., published in the Journal of Animal Science assessed the accuracy of genomic prediction in a large population of broiler chickens. The authors hypothesized that since the population of birds was large and genotyped, the accuracy of prediction would be high, and would not decay over time.
Hidalgo et al used predictive ability to assess the accuracy of genomic prediction to assess the accuracy of genomic selection over time, and determine if genotypic data collected from distant generations can be effectively used to predict genotypic and phenotypic responses to selection. Generally, the results from this study demonstrated that the accuracy of breeding value estimates for growth and production-related traits increased when more genotypic and phenotypic data were included in the analysis, but remained stable or only increased modestly when only phenotypic data were included for subsequent generations. Overall, the greater inclusion of recent, genotypic data into an analysis or estimate, the more accurate the estimates of the response to genetic selection.