Speaker
Description
Classical mean-field models of cell population dynamics, including cancer progression, often rely on the "well-mixed" assumption and neglect the role of cell death. In this work, we demonstrate how these simplifications lead to a significant overestimation of growth rates and saturation levels, with critical implications for tumor therapy design.
First, we present a stochastic modeling framework comparing mean-field approximations with Agent-Based Models (ABM) to investigate the "Ghost Effect" —a phenomenon where apoptotic cells do not vanish instantly but act as static barriers. These structures preserve spatial "memories" of former clusters, persistently inhibiting the motility and proliferation of living cells. Our results show that spatial correlations and volume exclusion consistently predict a lower carrying capacity than classical models.
Then, to bridge the gap between computationally expensive simulations and overly simplistic ODEs, we use a Moment Dynamics approach. By tracking higher-order moments, we derive a system of coupled ODEs that captures crowding effects and correlation-induced inhibition at a fraction of the computational cost.
Finally, we discuss applications in clinical oncology, highlighting how realistic modeling can help determine minimal effective chemotherapy doses by accounting for self-crowding.
