Turing patterns on discrete topologies: from networks to higher-order structures

Riccardo Muolo, Lorenzo Giambagli, Hiroya Nakao, Duccio Fanelli, Timotéo Carletti

TL;DR

This paper reviews the evolution of Turing’s theory of pattern formation from its classical roots in continuous partial differential equations (PDEs) to its modern application on complex networks and higher-order structures like hypergraphs. The authors demonstrate that discrete topologies are not just approximations of reality but powerful frameworks that enable "topology-driven" instabilities and patterns—such as oscillations in two-species systems—that are mathematically impossible in traditional continuous settings.

What Problem Does This Paper Address?

In 1952, Alan Turing proposed that a stable system of reacting and diffusing chemicals could spontaneously break symmetry to form patterns (like spots or stripes) if one species diffuses much faster than the other. While groundbreaking, classical Turing theory relies on continuous space and PDEs, which often come with very restrictive conditions that don't always match the abundance of patterns seen in nature.

The real world is often intrinsically discrete—think of neurons connected by axons, or species moving between discrete ecological patches. This paper addresses the need to generalize Turing’s ideas to these discrete topologies. By moving beyond simple grids to complex networks and "higher-order" structures (where interactions happen in groups rather than just pairs), researchers can model more realistic systems and discover new ways that patterns emerge.

Methods

The paper synthesizes several mathematical frameworks used to study dynamics on discrete supports:

  • Graph Laplacians: Replacing the continuous diffusion operator ($\nabla^2$) with a matrix representation of a network's connectivity.
  • Linear Stability Analysis: A standard technique to determine if a small perturbation will grow into a pattern, adapted here for the eigenvalues of the Graph Laplacian.
  • Master Stability Function (MSF): A tool borrowed from synchronization theory to analyze Turing-like instabilities in oscillatory or chaotic systems.
  • Higher-Order Tensors & Simplicial Complexes: Moving beyond pairwise links (A connects to B) to hyperedges (A, B, and C interact simultaneously).
  • The Dirac Operator: A novel approach used to couple "topological signals"—species that live on nodes (0-simplices) with species that live on links (1-simplices) or triangles (2-simplices).

Datasets

As a review paper focusing on theoretical frameworks, it does not introduce a new empirical dataset. Instead, it utilizes well-known mathematical models to demonstrate theory:

  • The Brusselator Model: A classic theoretical chemical reaction used to simulate pattern formation.
  • Synthetic Network Topologies: Simulations are performed on Small-World networks, Newman-Watts networks, and regular lattices (like triangular tilings) to visualize how different connection patterns influence the resulting "spots" or "oscillations."

Key Insights and Ideas

  • Directedness as a Catalyst: In continuous space, diffusion is usually symmetric. On directed networks (where A influences B but not vice versa), the resulting "non-normal" dynamics make it much easier for patterns to form. Crucially, directedness allows for oscillatory patterns in systems with only two species—a phenomenon that requires at least three species in the classical continuous recipe.
  • Non-Normality & Transient Growth: Even if a system is technically "stable" in the long run, the asymmetric nature of some networks can cause a massive "transient" spike in activity that triggers a pattern before the system can settle.
  • Patterns on the Links: Traditionally, we think of patterns as concentrations at specific points. This paper highlights "topological signals," where the pattern might manifest as a specific distribution of flows across the links of a network rather than just values at the nodes.
  • Temporal Ductility: For networks that change over time (temporal networks), the authors note that if the network reshuffles fast enough, the system behaves according to the "average" network. This suggests that high-frequency environmental changes might actually stabilize or enable certain biological patterns.

Key Results

  • The Topology-Driven Path: The authors highlight that the imaginary part of the Laplacian spectrum in directed networks acts as a "boost" for instability, effectively detaching the discrete dispersion relation from its continuous counterpart (see Figure 3).
  • Higher-Order Stabilization: While pairwise interactions might fail to produce a pattern, adding 3-body interactions (triangles) can either trigger an instability or, conversely, suppress one, providing a "tuning knob" for self-organization that doesn't exist in simpler models.
  • Dirac-Induced Patterns: By using the Dirac operator, the authors show that patterns can emerge from the coupling of different dimensions (e.g., a node-based activator and a link-based inhibitor), leading to a new class of "Dirac-induced" patterns.

Limitations and Future Work

  • Mapping Topology to Shape: While we know if a pattern will form, we still struggle to predict the exact shape of a pattern based solely on a complex network's architecture. A comprehensive "dictionary" between network metrics and pattern motifs is still missing.
  • Control Theory: The authors suggest that "pinning control"—manipulating a few key nodes to steer the entire system's pattern—is a promising but largely unexplored frontier for higher-order structures.
  • Quantum Frontiers: The paper briefly mentions the nascent study of Turing instabilities in quantum systems, suggesting that the discrete network approach provides a natural bridge to quantum frameworks.
  • Directed Higher-Order Structures: Most work on hypergraphs currently assumes symmetric interactions; extending this to directed hypergraphs is a major upcoming challenge.

Summary updated

Abstract

Nature is a blossoming of regular structures, signature of self-organization of the underlying microscopic interacting agents. Turing theory of pattern formation is one of the most studied mechanisms to address such phenomena and has been applied to a widespread gallery of disciplines. Turing himself used a spatial discretization of the hosting support to eventually deal with a set of ODEs. Such an idea contained the seeds of the theory on discrete support, which has been fully acknowledged with the birth of network science in the early 2000s. This approach allows us to tackle several settings not displaying a trivial continuous embedding, such as multiplex, temporal networks and, recently, higher-order structures. This line of research has been mostly confined within the network science community, despite its inherent potential to transcend the conventional boundaries of the PDE-based approach to Turing patterns. Moreover, network topology allows for novel dynamics to be generated via a universal formalism that can be readily extended to account for higher-order structures. The interplay between continuous and discrete settings can pave the way for further developments in the field.
BibTeX
@inproceedings{muolo2024turing,
  author = {Riccardo Muolo and Lorenzo Giambagli and Hiroya Nakao and Duccio Fanelli and Timotéo Carletti},
  title = {{Turing patterns on discrete topologies: from networks to higher-order structures}},
  booktitle = {Proceedings of the Royal Society A Mathematical Physical and Engineering Sciences},
  year = {2024},
  doi = {10.1098/rspa.2024.0235},
}

Loading related content...

Figure