Fig. 8
Integrative analysis validation. a Average adjusted R2 values for the training data sets (black bars) and the validation data sets (blue bars) for a five-fold cross-validation. Error bars denote the standard error of five sets of validations. b Adjusted R2 values for the lignin and wood properties predicted using either absolute transcript abundances as predictors (red bars) or metabolic-fluxes and metabolite concentrations as predictors (black bars). For both regression equations, an all-possible-model algorithm was used to determine the specific predictor variables. See Methods for detail. Adjusted R2 values represent variation explained by the regressions after adjusting for a different number of predictors
