The performance of various genome-based trait prediction methods applied to flowering time in multiple environments.

The best prediction model for each environment in the data set is in bold. The performance was evaluated via 10 runs of 10-fold cross-validation and the prediction accuracy was the mean Pearson correlation coefficients of predicted and observed FT.Finally, for further association study, genomic estimated breeding values (GEBVs) were predicted for each geographic region using RKHS with year as a covariate, respectively. RKHS implemented in the R package ‘rrBLUP’ is capable of handling multiple environments.The performance of various genome-based trait prediction methods applied to flowering time in multiple environments.