Development of a multivariate approach to predict Direct Genomic Values in dairy and beef cattle

The huge number of markers in comparison with the phenotypes available represents one of the main issues in genomic selection. In the present work Principal Component Analysis has been used to reduce the dimensionality of predictors. The method has been tested on three Italian cattle breeds (Holstein, Brown and Simmental) with different production aptitudes, dairy and dual purpose, and population size. Bulls were genotyped with the 54K Illumina beadchip and a data editing has been carried out. Remaining SNP have been used for further analysis. PC extraction was carried out separately for each chromosome and new variables able to explain different percentages of total variance were obtained. Bulls were sorted by birth year or randomly shuffled to create reference and prediction populations. The effect of principal components on polygenic EBV or Deregressed Proof in reference animals was estimated with a BLUP model. Results were compared to those obtained by using SNP genotypes as predictors either with BLUP or Bayes_A estimation methods. No substantial differences in correlations between DGV and EBV (or Deregressed Proof) were observed between the three methods in all experimental contributions. The PC method allows for a relevant reduction in the number of independent variables when predicting DGV, with a huge decrease in calculation time and without losses in accuracy.