Figure 1

PCA plots of microarray data from chemostat cultivations. Panel 1a ): All strain–carbon source datapoints, PC1 (principal component 1, x axis, 85.1% explained variation) vs. PC2 (y axis, 9.3% explained variation). Panel 1b ): All strain–carbon source datapoints, PC3 (x axis, 5.5% explained variation) vs. PC2 (y axis, 9.3% explained variation). Panel 1a and 1b will combine to show all three dimension in three first PCs. Panel 1c ): All datapoints from fructose-grown cultivations, PC1 (90.1% explained variation) vs. PC2 (7.4% explained variation). Squares, carbon source fructose; triangles, carbon source glycerol. Strains: Grey, WT; yellow, ΔalgC; red, mucA; blue, mucA ΔalgC; green, mucA TTalgD. All datapoints are independent fermentations, i.e. all biological conditions are present in triplicate data points. Genes used to construct the PCA are selected by ANOVA with p-values < 0.005.