Fig. 4: Dynamic phenotypes of primary metabolism predicted via the biosynthetic rates agree with experimental observations.

a, Yields of extracellularly exchanged metabolites with respect to glucose agree between cell-cycle-averaged flux predictions and independent measurements in an exponentially growing culture (x axis, mean ± s.d. from elsewhere⁸⁹; y axis data described below). b, Cell-cycle dynamics of predicted fluxes in the primary carbon and energy metabolism. c, Predicted turnover of cytoplasmic ATP during the cell cycle and the ATP fluxes in reactions that are largest producers or consumers of this metabolite. The turnover was calculated as the sum of ATP fluxes in reactions where this metabolite is produced. We show reactions whose cytoplasmic ATP flux is bigger than 0.09 or smaller than −0.09 mol cell⁻¹ h⁻¹ in at least one cell-cycle phase. PGK, phosphoglycerate kinase; PPCK, phosphoenolpyruvate carboxylase kinase; ADP/ATP, adenine nucleotide translocator (oxidative phosphorylation); PFK, phosphofructokinase; HEX, hexokinase. d, Predicted fluxes of biomass precursors diverting from central carbon and energy metabolism to the synthesis of major biomass components. NA, nucleic acids; PS, polysaccharides; e4p, erythrose 4-phosphate; pep, phosphoenolpyruvate; pyr, pyruvate; g6p, glucose 6-phosphate; f6p, fructose 6-phosphate; accoa, acetyl-CoA; glyc3p, glycerol 3-phosphate; r5p, ribose 5-phosphate. Vertical lines denote typical cell-cycle phases of major cell-cycle events (b–d). For presentation of data (y axis) in a–d: predictions shown by markers in a, line-connected bigger markers in b and c, heat map in c and lines in d correspond to the output of the cell-mass model provided with averaged replicate measurements of macromolecule biosynthesis (solid lines in Fig. 3b–d); predictions shown by y axis error bars (min–max range) in a and smaller markers in b and c correspond to the output of the cell-mass model using eight different combinations of replicate measurements (shaded area in Fig. 3b–d). e, Acquired intracellular fluorescence after a pulse of the glucose analog 2-NBDG varies during the cell cycle. Solid curve and shaded area indicate posterior mean and region of high posterior probability (mean ± s.d.) of the Gaussian process regression summarizing single-cell values (markers) via an RBF kernel. Dashed curve: posterior mean obtained via the same data analysis pipeline in the replicate experiment (number of analyzed cells indicated). f, Production rates of YFP and mCherry, having and lacking glycolytic flux regulation, respectively are uncoupled during the cell cycle in cells expressing the glycolytic flux biosensor. The uncoupling was calculated in individual cell‐cycle traces as the difference between the momentary production rates of YFP and mCherry normalized to have the same scale. A higher value of the uncoupling reflects a higher production rate of YFP with respect to the production rate of mCherry and thus a higher value of the glycolytic flux. Curve and shaded area show median and its 95% CIs. To align individual cell‐cycle traces and calculate phases, we used as reference points cytokinesis (CYT, 0), budding and next cytokinesis (CYT, 1).