When multi-trait data are available and the degree of correlation between phenotypic traits is moderate or large, genomic prediction accuracy can increase when models are used that account for correlations between the phenotypic traits. The data files associated with this dataset were used to explore Bayesian multi-trait kernel methods for genomic prediction. Linear, Gaussian, polynomial and sigmoid kernels were studied and compared with the GBLUP multi-trait models. The results of the analysis are reported in the accompanying article.
