For some applications of genomic prediction, it is appealing to combine information from multiple populations, an approach known as multi-population genomic prediction. Recently, a prediction equation was derived to predict the accuracy of genomic breeding values using this approach.
Multi-population genomic selection
Genomic information has the potential to increase the accuracy of breeding values, provided that large reference populations containing animals with both genotype and phenotype information are available. In numerical small breeds or populations, establishing such large reference populations is simply impossible. One way to increase the size of the reference population for numerically small populations is to add individuals from other populations to the reference population, for example individuals from different countries, breeds, or lines. The suitability of individuals from another population to increase the accuracy of genomic prediction is, however, reduced by differences between the populations.
In a recently accepted paper in Genetics, a prediction equation was derived to predict the accuracy of multi-population genomic prediction, accounting for the differences between the populations. The equation was shown to be able to accurately predict the accuracy of different multi-population genomic prediction scenarios. The equation is using the following input parameters; heritability of the trait, genetic correlations between populations, a measure for the relatedness between the populations and the proportion of the genetic variance captured by the markers. The derived prediction equation can be used to investigate the potential accuracy when populations from different lines, breeds or countries are combined in one reference population, or when populations measured for different traits are combined. Therefore, the equation is highly relevant to predict the benefit of combining populations in one reference population and can be used to optimize breeding programs.
For more information, please see the full article.