13.11 From Network Structure to System Dynamics

Throughout this chapter, we have represented biological systems as networks. This perspective has allowed us to identify hubs, modules, communication pathways, and recurring regulatory motifs. Network analysis has revealed that biological systems are highly organized rather than randomly assembled and that this organization reflects fundamental biological principles shaped by evolution.

However, an important question remains unanswered.

Knowing who interacts with whom does not necessarily explain what the system actually does.

Learning objectives

After studying this chapter, you should be able to

explain the limitations of purely structural network analysis,
distinguish between network topology and system dynamics,
understand why biological interactions require quantitative models,
appreciate the role of dynamical systems in systems biology,
understand the transition from network biology to mathematical modelling.

Networks describe possibilities

A network tells us which interactions are possible.

If two proteins are connected in a protein interaction network, they are capable of interacting.

If two metabolites are connected in a metabolic network, one can be converted into the other.

If a transcription factor regulates a gene, information can flow between them.

Network topology therefore defines the architecture of the system.

It answers questions such as

Which components are connected?
Which pathways exist?
Which proteins act as hubs?
Which modules cooperate?

These are important questions because network architecture constrains the behaviour of the system.

Nevertheless, topology alone does not determine behaviour.

The same network can behave very differently

Consider a simple regulatory interaction.

Gene A activates gene B.

This relationship can produce many different behaviours.

Gene B may respond immediately.

It may respond only after a delay.

It may react weakly or strongly.

It may activate only above a certain threshold.

It may even oscillate if additional regulatory mechanisms are present.

The network itself does not provide this information.

The graph tells us that an interaction exists.

It does not tell us how strongly, how rapidly, or under which conditions the interaction occurs.

The behaviour of the system therefore depends not only on the network structure but also on the dynamics of its interactions.

Biological systems change over time

Living systems are dynamic.

Gene expression changes continuously.

Protein concentrations increase and decrease.

Metabolites are produced and consumed.

Cells respond to environmental signals, adapt to stress, divide, differentiate, and eventually die.

Consequently, the state of a biological system is never fixed.

Instead, it evolves over time.

Describing these temporal changes requires a different type of model than static network representations.

From topology to dynamics

The transition from network biology to dynamic systems can be understood through a simple analogy.

A road map describes which cities are connected by roads.

However, it does not describe traffic.

Traffic depends on

the number of vehicles,
driving speed,
traffic lights,
road conditions,
time of day.

Similarly, a biological network specifies which molecular interactions are possible.

The actual behaviour depends on

molecular concentrations,
reaction rates,
regulatory strengths,
delays,
nonlinear responses,
stochastic fluctuations.

Network topology provides the framework.

Dynamics determines how the system behaves within that framework.

Why mathematical models are needed

To understand biological dynamics, we must move beyond static graphs.

Instead of asking

Which components interact?

we now ask

How does the state of the system change over time?

Answering this question requires mathematical models capable of describing continuous change.

Depending on the biological problem, these models may take many forms.

Examples include

ordinary differential equations,
stochastic models,
Boolean networks,
agent-based models,
hybrid models.

All share a common objective:

to describe how biological systems evolve over time.

The next step in systems biology

Network biology provides the structural foundation of systems biology.

It identifies the components of a system, reconstructs their interactions, and reveals the organizational principles governing biological networks.

Dynamic systems biology builds upon this foundation.

Rather than studying the architecture alone, it investigates how biological states change, how regulatory circuits generate behaviour, and how complex dynamics emerge from interacting components.

Together, network biology and dynamic modelling provide complementary perspectives on living systems.

One tells us how the system is organized.

The other explains how the system behaves.

Take-home message

Networks describe the architecture of biological systems. Dynamic models explain how biological behaviour emerges from that architecture.

Key concepts

Networks describe interaction structure but not system behaviour.
Biological systems are dynamic and continuously change over time.
Network topology constrains, but does not determine, biological dynamics.
Mathematical models are required to describe temporal changes.
Dynamic systems biology extends network biology by modelling biological behaviour over time.

Summary

Network biology provides a powerful framework for representing and analysing the interaction structure of biological systems. However, structure alone is insufficient to explain biological behaviour. Living systems continuously change over time, and understanding these changes requires dynamic mathematical models. The next part of this book introduces these models and explores how biological behaviour emerges from interacting molecular networks.

Self-check questions

What information is contained in a biological network?
Why is network topology alone insufficient to predict biological behaviour?
Why can identical network structures exhibit different dynamics?
Which biological quantities change over time?
Why are mathematical models required to study biological dynamics?
How do network biology and dynamic systems biology complement one another?