Figure 2
Cross-validated accuracies of prediction ( r p ) for genomic selection using RR-BLUP across genotype (CV G ). Accuracies of prediction were determined within and across populations. ES in the schemes on top refers to the estimation set, TS represents the test set. In each case, the dataset was divided into five subsets (S1-S5), of which S1 to S4 were assigned to the ES and S5 to the TS. Box-whisker plots of accuracy of prediction rp values for traits grain yield, plant height, starch content, and total pentosan content based on data from population A (Pop-A) and population B (Pop-B) follow below. Cross-validation was performed with estimation and test sets derived from one population, either Pop-A or Pop-B (CVG Within-Within-Same; white; category I), estimation and test sets derived from both populations, Pop-A and Pop-B (CVG Across-Across; dark grey; category II), the estimation set derived from both populations, Pop-A and Pop-B, and the test set from one population, either Pop-A or Pop-B (CVG Across-Within; light grey; category III), or estimation set derived from one population, either Pop-A or Pop-B, and the test set from one population, either Pop-B or Pop-A (CVG Within-Within-Different; intermediate grey; category IV).