Fig. 3From: Genomic dissection of additive and non-additive genetic effects and genomic prediction in an open-pollinated family test of Japanese larchDistribution of predictive abilities for each trait. Each panel contains box plots showing the distribution of predictive abilities from the base models (GBLUP-A), and the optimal models (without genetic terms that provide a negligible contribution to the phenotypic variance) considering non-additive effects fitted by the GBLUP (GS-GBLUP) and RKHS (GS-RKHS) approaches. ns represents non-significance; * and ** indicate significant at p = 0.05 and p = 0.01. Hollow circles represent outliers. The black solid points and bars represent means and medians. See Table 1 for full description of traitsBack to article page