A Two-Stage Method for Improving the Prediction Accuracy of Complex Traits by Incorporating Genotype by Environment Interactions in Brassica napus
Table 5
The prediction accuracy (standard deviation, SD) of four methods for SY.
Environment
LASSO
OLS post-LASSO
Stratified GBLUP
M × E GBLUP
E7
0.4952 (0.0082)
0.5735 (0.0075)
0.4418 (0.0076)
0.4767 (0.0071)
N3
0.1600 (0.0104)
0.6958 (0.0092)
0.1336 (0.0097)
0.2322 (0.0104)
N4
0.2635 (0.0115)
0.3816 (0.0089)
0.1509 (0.0110)
0.3049 (0.0098)
N6
0.2811 (0.0097)
0.7039 (0.0092)
0.4038 (0.0089)
0.4653 (0.0097)
N7
0.2224 (0.0106)
0.6312 (0.0089)
0.3251 (0.0079)
0.4464 (0.0087)
S3
0.1351 (0.0111)
0.4698 (0.0108)
0.0907 (0.0110)
0.2088 (0.0114)
S4
0.2887 (0.0109)
0.6395 (0.0082)
0.3056 (0.0099)
0.4269 (0.0094)
S5
0.1232 (0.0093)
0.2285 (0.0113)
0.3595 (0.0106)
0.3924 (0.0079)
S6
0.4142 (0.0100)
0.6203 (0.0074)
0.3973 (0.0099)
0.5919 (0.0070)
S7
0.3351 (0.0093)
0.5627 (0.0092)
0.2517 (0.0099)
0.4597 (0.0087)
Average
0.2718 (0.0101)
0.5507 (0.0091)
0.2860 (0.0096)
0.4005 (0.0090)
SY, seed yield per plant; LASSO, least absolute shrinkage and selection operator; OLS, ordinary least squares; GBLUP, genomic best linear unbiased prediction; M × E, marker × environment interaction.
