20.1 Why Deterministic Models Are Not Enough

20.1 Why Deterministic Models Are Not Enough

Throughout this book, we have described biological systems using deterministic models. Ordinary differential equations (ODEs) have allowed us to understand regulatory networks, metabolic pathways, population dynamics, oscillations, and bifurcations. In all these examples, the state of a system changes continuously according to well-defined mathematical rules. Once the initial conditions and model parameters are known, the future trajectory of the system is uniquely determined.

This deterministic perspective has proven extraordinarily successful and remains one of the cornerstones of systems biology. However, it relies on an important assumption: the quantities being modeled are sufficiently large that individual molecular events average out. In other words, fluctuations caused by single reactions are assumed to be negligible compared to the overall dynamics.

For many biological processes, this assumption is entirely justified. Cellular metabolites may be present in millions of copies, and metabolic fluxes often involve enormous numbers of reaction events. Under these conditions, random fluctuations become insignificant, and average concentrations provide an accurate description of the system.

Gene regulation presents a fundamentally different situation.

Many regulatory molecules exist in only a handful of copies within a cell. A bacterial gene may have only a single DNA copy, a few messenger RNA molecules, and perhaps only tens of copies of a regulatory protein. At these small copy numbers, every molecular event matters. The binding of a single transcription factor, the degradation of a single mRNA molecule, or the initiation of a single translation event can substantially alter the future behavior of the system.

As a consequence, two genetically identical cells exposed to exactly the same environmental conditions may follow different molecular trajectories. One cell may produce a burst of protein expression while another remains almost silent. These differences are not caused by hidden variables or experimental errors—they arise naturally from the probabilistic nature of molecular interactions.

This observation forces us to reconsider what a mathematical model should predict. In deterministic models, we seek the unique trajectory of the system. In stochastic systems, however, there is no single trajectory that represents all cells. Instead, every realization is only one possible outcome among many equally plausible alternatives.

The natural question therefore changes.

Instead of asking,

What will happen?

we begin to ask,

What is the probability that a particular event or cellular state will occur?

This seemingly small change in perspective has profound consequences. The object of interest is no longer a single trajectory but an entire probability distribution describing all possible states of the system.

The Chemical Master Equation provides exactly this description. Rather than predicting the future state of one individual cell, it predicts how the probabilities of all possible cellular states evolve over time.

Before introducing this mathematical framework, however, we first need to understand where stochasticity originates in living cells and why biological systems exhibit such remarkable variability despite being genetically identical.