The agent-based model is validated by experimental data. (A) Growth characteristics of co-cultured human mammary epithelial cells (HMEC) (labeled with vital dye Qtracker565; Invitrogen) and variant human mammary epithelial cells (vHMEC) (unlabeled) were recorded by time-lapse microscopy. An example is presented of the changes in proportions of cells present in a field of view monitored over seven days. (B) Growth of co-cultured HMEC and vHMEC was compared with ABM simulations. Each data point represents the mean of four biological replicates. Pre-stasis HMEC (red diamonds) stopped proliferating after four days in culture in contrast to vHMEC (green triangles) that continued to proliferate until the cultures achieved confluence. Solid lines show agent-based simulations, based on the initial cell densities, cell cycle duration during the proliferative phase (26 hours for both cell types), maximum cell compression observed at full confluence, and an intrinsic limit on the number of population doublings achievable by HMEC. Note that the simulations of the model were well correlated with the experimental data, indicating that kinetic growth rates could be predicted accurately for both cell types using a limited number of assumptions. (C) Illustration of growth simulation. Normal growing HMEC (red) undergo limited divisions before losing proliferative potential and undergoing stasis (blue). vHMEC (green) are capable of unlimited of divisions in this example. At low density (Day 1), most cells occupy a maximum area (that is, 21 pixels equivalent to 175 μm2, matching light microscopy measurements). Cells continue dividing as space becomes more restricted (Day 5), until the available adjacent space reaches a minimum of six pixels (50 μm2 at Day 14). Such simulations accurately recapitulate contact inhibition as measured by time-lapse microscopy.