Skip to main content
. 2016 Oct 28;7(1):41–53. doi: 10.1534/g3.116.035584

Table B3. Wheat data set 3.

Env. Covariance Matrix UE (Upper Triangular) and Correlation Matrix (Lower Triangular) for u Covariance Matrix FE (Upper Triangular) and Correlation Matrix (Lower Triangular) for f Variance–Covariance Matrix Σ for ε
1 2 3 4 1 2 3 4 1 2 3 4
GBLUP
 E1 0.403 0.368 0.129 0.169 0.254 0.086 0.087 0.017 0.281
 E2 0.729 0.632 0.329 0.204 0.403 0.179 0.036 0.067 0.184
 E3 0.273 0.555 0.556 0.128 0.362 0.178 0.228 0.033 0.322
 E4 0.448 0.432 0.289 0.353 0.070 0.327 0.143 0.234 0.366
GK
 E1 0.693 0.453 0.175 0.191 0.132 −0.014 0.033 −0.023 0.145
 E2 0.638 0.727 0.302 0.248 −0.122 0.099 0.004 0.012 0.123
 E3 0.229 0.386 0.841 0.171 0.267 0.037 0.116 −0.004 0.126
 E4 0.269 0.342 0.219 0.725 −0.168 0.101 −0.031 0.142 0.163

Empirical phenotypic correlation: E1 vs. E2 = 0.527; E1 vs. E3 = 0.253; E1 vs. E4 = 0.259; E2 vs. E3 = 0.340; E2 vs. E4 = 0.328; E3 vs. E4 = 0.220. Variance–covariance matrix (upper triangular) and correlation matrix (lower triangular) for random effects u, f, and variance matrix for random errors ε of multi-environment model (3) including four environments (E1–E4) for linear kernel GBLUP and nonlinear Gaussian kernel (GK). Pair-wise sample phenotypic correlations between environments are given above. Env., environment; GBLUP, genomic best linear unbiased predictors: GK, Gaussian kernel.