Single cell statistics of the wild type strain (mass doubling time min): experimental and simulation values (before and after parameter optimization).

Parameter optimization using the experimental data in Table 1 does not affect the overall fitting error in terms of wild type statistics (these data are not enforced during optimization). Fitting error is defined as , where and are the statistical data points in simulations and experiments, respectively (: mean cycle time of mothers, : mean cycle time of daughters, : mean G1 duration of mothers, : mean G1 duration of daughters, : CV of cycle time among mothers, : CV of cycle time among daughters, : CV of G1 duration among mothers, : CV of G1 duration among daughters, and through denote the simulation values for the same statistics). Simulation statistics (mean standard deviation) are computed from 15 independent realizations. In each realization, eight pedigrees are generated. Each pedigree of cells is initiated by a single daughter (D) or mother (M) cell. CV denotes coefficient of variation (standard deviation normalized by the mean), whereas represents the G1 duration. Experimental mass doubling time of 100 minutes [10] is used in the simulations. The number of failed cycles (due to event execution errors listed in Table S9) normalized by the number of complete cycles is the cycle failure ratio. Matlab script to reproduce the mean and CV values (rightmost column) is provided as File B in File S1.