Interpretive Summary: Genomic selection in American mink using a ssGBLUP model
By: Anne Kamiya, MS
Fertile minks with large, high quality pelts are desired in the mink farming industry. Heritability of desired traits can range from low to high, are not necessarily inherited together and can be difficult to predict based on pedigree. For instance, although female minks that produce large litters are desired, minks with higher body weights (larger pelts) also tend to be less fertile.
In livestock animals, genomic predictability models are more reliable than pedigree-based models when selecting for desired traits. The single-step genomic best linear unbiased prediction (ssGBLUP) model, for instance, is highly accurate as it combines data obtained from genotype, phenotype and pedigree to determine estimated breeding values (EBVs). For the genomic portion of this model, single-nucleotide polymorphisms (SNPs) – point mutations in a single nucleotide which act as genetic markers for identifying desired traits – are normally used. However, in mink SNP data are lacking. Genotyping by sequencing (GBS) is an alternative to SNP for obtaining genomic markers.
In this recent Journal of Animal Science study, the authors evaluated the usefulness of GBS in American mink, comparing the predictive accuracy of the standard pedigree BLUP model versus the ssGBLUP model. To achieve this goal, EBVs from pedigree BLUP models were compared against ssGBLUP models with GBS data. The ssGBLUP model was found to be more accurate than the pedigree model for body weight and quality of pelt and also had less EBV bias.
Overall, the results of this study suggest that using GBS data in place of SNP (when SNP is lacking) is likely sufficient in mink and that a ssGBLUP model is more accurate and less biased than the BLUP model. Further work into developing ssGBLUP models to replace BLUP models in mink and other animals is justified.The full paper can be found on the Journal of Animal Science webpage.