17.9 Modelling Lessons: Constructing Models of Biological Oscillators

Throughout this chapter, we have studied several biological oscillators that differ enormously in their molecular components and physiological roles. Hormone regulation, respiratory control, motor control, and gene expression appear to be unrelated biological systems.

Yet, from the perspective of dynamical systems, they share a remarkably similar mathematical structure.

This illustrates one of the central ideas of systems biology:

The purpose of mathematical modelling is not to reproduce every biological detail, but to identify the dynamical principles that govern system behavior.

When constructing a model of an oscillatory biological system, it is therefore useful to follow a systematic modelling strategy.

Step 1: Identify the feedback architecture

Oscillations almost always arise from feedback.

The first question should therefore never be:

Which equation should I use?

Instead, ask:

Which components regulate each other?

Construct a qualitative interaction diagram before writing down any equations.

Identify

positive regulations,
inhibitory interactions,
and closed feedback loops.

Only after understanding the architecture of the network should mathematical equations be introduced.

Step 2: Decide what the variables represent

Every mathematical variable represents a biological quantity.

Before writing a model, decide carefully what each variable describes.

Examples include

mRNA concentrations,
protein concentrations,
metabolite levels,
hormone concentrations,
population sizes,
or physiological quantities such as body temperature.

Choosing appropriate state variables determines the level of abstraction of the model.

A simple model often combines many molecular components into a single variable representing an entire biological process.

The goal is not maximum detail but an appropriate level of complexity for the biological question being addressed.

Step 3: Translate biological knowledge into mathematical functions

One of the most important modelling principles developed throughout this book is that mathematical functions are not chosen arbitrarily.

Instead, they encode biological assumptions.

Before selecting a function, ask:

What qualitative behavior should this biological process exhibit?

For example,

linear functions describe proportional relationships,
sigmoidal functions describe threshold-like responses,
Hill functions describe cooperative regulation and saturation,
degradation terms represent continuous turnover,
delay terms represent transport or processing times.

Different biological assumptions naturally lead to different mathematical functions.

The function is therefore a formal representation of biological knowledge.

Step 4: Identify nonlinearities

Linear systems are mathematically convenient, but many biological phenomena arise precisely because biological regulation is nonlinear.

Examples include

enzyme saturation,
cooperative binding,
transcriptional activation,
all-or-none switching,
and threshold-dependent responses.

These nonlinearities often determine whether a system exhibits homeostasis, switching behavior, or oscillations.

Whenever a biological process appears highly sensitive or saturates at large concentrations, a nonlinear function should be considered.

Step 5: Look for sources of delay

Real biological regulation is rarely instantaneous.

Information requires time to propagate through biological systems.

Possible sources of delay include

molecular transport,
transcription,
translation,
protein folding,
intracellular trafficking,
neural signal transmission,
hormonal circulation,
or developmental processes.

Even relatively small delays can fundamentally alter system dynamics.

Whenever information must travel through multiple intermediate steps, delays should be considered explicitly during model construction.

Step 6: Identify potential control parameters

Finally, ask which parameters might qualitatively change the behavior of the system.

Many parameters simply modify numerical values without changing the overall dynamics.

Others fundamentally alter the operating mode of the system.

Typical control parameters include

feedback sensitivity,
transport delays,
degradation rates,
interaction strengths,
or external inputs.

Exploring these parameters systematically often reveals bifurcations that would be impossible to discover by intuition alone.

This is one of the greatest strengths of mathematical modelling.

A modelling recipe for biological oscillators

The examples in this chapter can be summarized by the following general workflow.

Identify the regulatory network.
Determine the state variables.
Translate biological interactions into mathematical functions.
Include nonlinear regulation where biologically justified.
Consider explicit or implicit delays.
Identify control parameters and explore their effects on system dynamics.
Analyze the resulting model to determine whether it predicts stable equilibria, oscillations, or bifurcations.

Notice that mathematical analysis comes after biological reasoning.

The equations do not replace biological understanding.

They provide a rigorous language for expressing biological hypotheses and testing their dynamical consequences.

Looking ahead

This chapter introduced oscillations as a second fundamental type of biological equilibrium.

In the previous chapter, we learned how nonlinear systems can generate multiple stable states and biological switches.

Here we have seen that the same modelling framework naturally explains rhythmic biological behavior through stable limit cycles.

Together, these concepts illustrate an important message that will continue throughout this book:

The behavior of biological systems emerges from the interaction between network architecture and dynamics.

Mathematical models allow us to move beyond describing biological observations and begin explaining why biological systems behave as they do.

Key concepts

Begin modelling with the biological interaction network rather than with equations.
Mathematical functions encode biological assumptions.
Nonlinearities often determine qualitative system behavior.
Delays are natural consequences of biological information processing.
Control parameters determine whether a system exhibits homeostasis, switching, or oscillations.
Mathematical models reveal dynamical principles that are difficult to recognize experimentally.

Summary

Building models of biological oscillators requires much more than writing differential equations. Effective models begin with biological knowledge, identify the underlying regulatory architecture, and translate qualitative biological assumptions into mathematical functions. Nonlinear regulation and time delays frequently determine whether a system exhibits stable homeostasis or sustained oscillations. By systematically exploring control parameters, mathematical models reveal the dynamical principles that organize biological behavior across many different scales.

Self-check questions

Why should model construction begin with the biological interaction network?
Why are mathematical functions chosen based on biological assumptions?
Give three examples of biological nonlinearities.
Why are delays common in biological regulation?
What distinguishes an ordinary parameter from a control parameter?
Why is parameter exploration an essential part of modelling?
Summarize the general workflow for constructing a model of a biological oscillator.