Fig. 9
Scatterplots of the phenotypic variations captured by the lignin-based integrative analysis. The monolignol transcript abundance was used as the only input for the multi-level integrative analysis to predict the corresponding protein abundance, predicted metabolic-flux, predicted metabolite concentrations, and the 25 lignin and wood properties. The figure shows that transcript abundance is used to predict the 25 lignin and wood properties (y-axis, through the multi-level analysis), which are regressed on the experimental data (x-axis, measured in transgenics and wildtype). a Lignin content; b S/G ratio; c S-subunits; d G-subunits; e H-subunits; f p-Hydroxybenzoic acid; g Carbohydrate to lignin (C:L) ratio; h Aldehyde content; i β–O–4 linkages; j β–5 linkages; k β–β linkages; l β–1 linkages; m Cinnamyl alcohol end-groups; n Height; o Diameter; p Stem volume; q Glucose content; r Xylose content; s Total carbohydrates; t Relative density; u MOE; v–y Saccharification efficiency for glucose (v, x) and xylose (w, y) release from unpretreated (v, w) or pretreated (x, y) wood samples. The R2 values are coefficients of determination, showing the percent variation in dependent variables explained by monolignol gene transcript abundance as the predictor. Each data point represents a replicate pool of one wildtype or transgenic line containing 3–5 clonally propagated trees
