Fig. 6: The interplay of inflation-selection balance and evolutionary rescue inherently emerges in an agent-based in silico tumor model.
a Schematics of the simulation set-up (see “Methods” section for details). b Simulated trajectories. Line width is proportional to the clone width at the colony edge. c Width of compensated (blue) and uncompensated (red) clones. Solid and dashed lines represent the median and mean widths, respectively, and shaded areas indicate interquartile ranges. s = 0.21 (see Supplementary Fig. 12 for different fitness costs). d, e Estimated probability densities (see Fig. 5d, e and “Methods” for details). f Clone survival probabilities at the front (Eq. (1)). Shaded area indicates Poisson distribution SD. g Efficacy of compensatory mutations. Shaded areas indicate propagated SDs. Gray box represents the window of inefficacy. h Representative simulated clone trajectory exhibiting all phases (see main text). i Mean clone width development of uncompensated clones (red lines) and mutations triggered at different colony radii r* (blue solid lines). Lines begin at their mutation time point. Hue decreases with the mutation start. Dashed lines are linear fits to the width of compensated clones. Shaded areas represent SD of the fitted lines. j Radial expansion Δrescape from the point of mutation r* to escape wescape as function of r* (mean ± SD). k Probability Pescape (black circles) to grow above escape width wescape = 6 cells in 2500 μm of radial colony growth after mutation. Error bars indicate one SD assuming a Poisson distribution. Probability Pmutate (black diamonds) to get a mutation (with mutation rate μ = 0.1 per step) at the given radius. Combined probability Ptotal (red squares) for a clone to acquire a mutation and eventually grow to the width w > wescape. n = 20 simulated colonies for each r*. Source data are provided as a Source data file.
