Fig. 2

Feasible region (error ≤ 0.05) identified using equilibrium fluorescence cytometry data and depicted according to three different characteristic plots. a Data fitting error (colour-coded) of the feasible region as a function of population-averaged production (\(\bar P\)) and degradation (\(\bar D\)) rates for N = 15. Numerals indicate areas for which representative results are shown in the accordingly numbered small images (display analogous to Fig. 1). The asterisk indicates the error minimum. b Population-averaged correlation as a function of the population-averaged slopes of production (\(\overline {{\mathrm d}P}\)) and degradation (\(\overline {{\mathrm d}D}\)) rates. c Average absolute Fokker–Planck-associated diffusion term as a function of the corresponding average absolute deterministic term \(\overline {|A_{{\mathrm {FP}}}|}\) and regulated noise term \(\overline {|DB_{\mathrm {FP}}|}\). d–f Analogous results for N = 60. g–j Results for the same data but assuming zero proliferation during modelling. Panels g, h relate to a, c, panels i, j to d, f. Colour in 2D plots corresponds to averages according to a 600 × 600 grid. Reduced growth rate 0.65·0.5·ln(2)/h (experimental measurement) in a–f and assumed zero growth in g–j. Population-averaged correlation is calculated first per cell cycle phase i weighting P(n, i) and D(n, i) by the probability p(n, i) and, second, by averaging across i (see Methods). Note that we increased the error limit to 0.07 for N = 60 in i, j because of reduced efficiency of the zero proliferation model. Original data by courtesy of A. Kashiwagi and T. Yomo