Table 7 Predictive abilities, per trait (EH, GY, and PH) and average across traits (μ), as training set size increases (OTS 1, 2, and 3), for GET (genotype x environment x trait) and GWT (genotype x environmental covariates x trait) kernels, for the HEL (Helix) and USP (University of Sao Paulo) datasets. Ne: number of information used as the training set; prediction accuracies for EH: ear height, GY: grain yield, and PH: plant height; %: the percentage increase between that value and the value immediately above; OTS: optimized training set. Here, Ne was added with the check’s information into the original sample sizes, being 9 information for HEL and 12 information for USP

From: Enviromic-based kernels may optimize resource allocation with multi-trait multi-environment genomic prediction for tropical Maize