Table 4 MAIZE3 data set. Average Pearson’s correlations between observed and predictive values (and their standard deviation in parentheses) for 3 methods for single-environment and GE multi-environment models for 50 random partitions with 70% of the lines in the training set and 30% of the lines in the testing set. Methods GB, GK, and AK are the GBLUP, Gaussian Kernel and Arc-Cosine Kernel, respectively. The best predictive model for each environment is in boldface
Single-environment model
EnviromentGBAKGK
E10.246 (0.05)0.271+ (0.05)0.273 (0.05)
E20.321 (0.04)0.319 (0.04)0.320 (0.04)
E30.295 (0.06)0.294 (0.06)0.294 (0.06)
E40.42 (0.05)0.421 (0.05)0.422 (0.05)
GE multi-environment model
E10.301 (0.05)0.520+ (0.05)0.517 (0.05)
E20.342 (0.04)0.505+ (0.05)0.497 (0.04)
E30.330 (0.05)0.478+ (0.04)0.470 (0.04)
E40.435 (0.05)0.555+* (0.04)0.546 (0.05)
  • * Significant at the 0.05 probability level of the t-test for the hypothesis that the average of the correlation of kernel AK is superior to the mean of the correlation of kernel GK.

  • +Significant at the 0.05 probability level of the t-test for the hypothesis that the average of the correlation of kernel AK is superior to the mean of the correlation of kernel GB.