Figure EV2. Growth rate fluctuations in representative microcolonies after TMP, TET, or NIT addition and their effect on timing variability determined with different response time measures.

1. Instantaneous growth rates (smoothed with moving average filter of window size 3) for each condition were determined for single cells (thin lines) in a microcolony, acquired with a microfluidics system (CellASIC). The thick lines represent the mean over all depicted single cells.
2. Same plot as in Fig 2C, but with the response time mean and standard deviations determined from the number of cell doublings instead of time. This analysis corrects for fluctuations in the growth rates of single cells. The dashed and dotted lines are the same as in Fig 2C.
3. Same data as in Fig 2A, but each cell was normalized to its maximum expression level (Materials and Methods). Response times in Fig EV2D–G were defined as the time until 50% of the maximum expression level was reached for each single cell.
4. Same plot as in Fig 2C, but with the response time determined as depicted in Fig EV2C.
5. Same plot as in Fig EV2B, but with the response time determined as depicted in Fig EV2C with cell doublings instead of time.
6. The randomness parameter (σ_t/μ_t)² compared for the two tested response time measures. X‐axis: Time until 25% of the median full expression of the microcolony was reached (see Fig 2A); y‐axis: Time until half maximal expression was reached for each cell individually (see Fig EV2C).
7. Same plot as in Fig EV2F, but with the response times determined in units of the cell doubling time.
8. Standard deviation divided by the mean response time over the maximal absolute expression level (obtained from population‐level measurements of the GFP‐promoter library; Mitosch et al, 2017) for individual promoters after the addition of TMP, TET, or NIT. The Pearson correlation coefficient is −0.08 ± 0.23, error and P‐value from bootstrapping (Materials and Methods).