We use agent-based models as a general approach to explore how the dynamics of complex systems emerge from interactions among their constituent parts. The systems we care about are cells, subcellular structures (e.g a patch of the cell cortex), and multicellular tissues. The parts - or agents - are molecules, macromolecular structures (e.g. filaments or organelles, etc), discrete pieces of cells (again like a patch of the cell cortex, or a piece of internal cytoplasm, or a mixture of all three. The agent-based models that we use encode, as systems of ordinary differential equations, empirically-based rules for how each part behaves and how it interacts with every other part. Numerical solution of the equations predicts the system-level behaviors that emerge from these local rules.
Agent-based models allow us to ask: Does what we know about the parts and their interactions add up to an explanation for the emergent dynamics of the whole system? If not, what might be missing? If so, can we infer simpler mechanistic principles that govern the behavior of the system? Can we design key experiments to test these principles?