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

Table B4. Wheat data set 4.

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 0.483 0.111 0.004 0.133 0.409 0.230 −0.055 0.215 0.175
 E2 0.234 0.467 0.291 0.246 0.585 0.378 0.091 0.199 0.242
 E3 0.008 0.560 0.578 0.292 −0.163 0.280 0.279 0.105 0.230
 E4 0.304 0.572 0.610 0.396 0.541 0.521 0.320 0.386 0.243
GK
 E1 0.968 0.353 −0.042 0.378 0.142 0.084 −0.021 0.059 0.111
 E2 0.395 0.826 0.417 0.476 0.502 0.197 −0.007 0.045 0.176
 E3 −0.044 0.470 0.952 0.439 −0.170 −0.048 0.107 0.009 0.129
 E4 0.429 0.584 0.502 0.803 0.383 0.248 0.067 0.167 0.192

Empirical phenotypic correlation: E1 vs. E2 = 0.342; E1 vs. E3 = −0.054; E1 vs. E4 = 0.311; E2 vs. E3 = 0.328; E2 vs. E4 = 0.414; E3 vs. E4 = 0.223. 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 above. Env., environment; GBLUP, genomic best linear unbiased predictors: GK, Gaussian kernel.