FIGURE 7.

Prediction accuracies across environment Lind with training on the Pullman dataset for seven end-use quality traits using four different uni- and multi-trait and one Bayesian multi-trait multi-environment genomic prediction models.