Relay transitions and invasion thresholds in multi-strain rumor models: a chemical reaction network approach

This paper applies the Chemical Reaction Network theory (CRNT) and the symbolic package EpidCRN to analyze multi-strain rumor spreading models on online social networks, revealing that stability transitions occur via "relay" mechanisms governed by siphon-induced transcritical bifurcations and invasion inequalities.

The Gist

Here is an explanation of the paper using simple language, everyday analogies, and creative metaphors.

The Big Picture: A New Way to Watch Rumors Spread

Imagine you are watching a rumor spread through a massive online social network. You have a computer model (a set of equations) trying to predict what will happen. Will the rumor die out? Will it take over? Will two different rumors fight each other?

Usually, mathematicians look at these models by checking specific "snapshots" (equilibrium points) to see if they are stable. If a snapshot is unstable, they ask: "What happens next?"

This paper introduces a new, more organized way to answer that question. The authors, using tools from Chemical Reaction Networks (how molecules react) and Mathematical Epidemiology (how diseases spread), propose that these systems don't just jump randomly from one state to another. Instead, they move in a structured, step-by-step chain called a "Relay."

Think of it like a baton pass in a relay race. The "baton" is the stability of the system. When one runner (a specific state of the network) gets tired (becomes unstable), they pass the baton to the next runner (a new state) who is waiting right next to them.

The Core Concepts, Explained Simply

1. The "Siphon" (The Trap Door)

In the world of chemical reactions and rumors, a Siphon is like a trap door or a one-way valve.

• The Metaphor: Imagine a room with a door that only opens out. If everyone leaves the room, no one can get back in. Once the room is empty, it stays empty forever.
• In the Paper: A "siphon" is a group of variables (like "users who believe Rumor 1") that, if they drop to zero, stay at zero. You can't spontaneously generate believers out of thin air; you need a source.
• Why it matters: These siphons create "walls" or "faces" in the mathematical model. The system is trapped inside these walls. The authors realized that the entire behavior of the rumor network is organized by a lattice (a grid-like structure) of these walls.

2. The "Relay" (The Baton Pass)

This is the paper's main discovery.

• The Metaphor: Imagine a game of musical chairs, but the chairs are arranged in a specific hierarchy. When the music stops, the person sitting in Chair A (the current state) realizes the chair is wobbly (unstable). They don't fall randomly; they step directly into Chair B, which is right next to them.
• The "Coincidence": The authors found something magical: The exact condition that makes Chair A wobbly (unstable) is the same condition that makes Chair B appear (stable).
  • If the "invasion number" (a measure of how strong the new rumor is) is high enough to knock out the current state, it is automatically high enough to create the new state.
  • It's like a light switch: The moment the switch flips off (old state dies), the light turns on (new state is born). There is no gap, no chaos, just a smooth handover.

3. The "Invasion Number" (The Rumor's Strength)

In disease models, we use $R_0$ (Basic Reproduction Number) to see if a virus spreads. Here, they use Invasion Numbers.

• The Metaphor: Think of a rumor as an invader trying to enter a fortress (the current network state).
• If the rumor is weak (Invasion Number < 1), it bounces off the walls. The fortress stays the same.
• If the rumor is strong (Invasion Number > 1), it breaks the gate. The fortress collapses, and a new fortress (a new state where the rumor exists) is built immediately on the ruins.

The Case Study: The Online Social Network (OSN) Model

The authors tested their theory on a specific model of how two rumors spread on a social network. The model has 8 different "compartments" (types of people):

• Potential users, new users, believers of Rumor 1, believers of Rumor 2, skeptics, and people who quit the platform.

They looked at two scenarios:

Scenario A: The Simple Case ($\omega = 0$)

• The Setup: People who become skeptics stay skeptics forever. They don't go back to believing or leave the platform.
• The Result: Everything is perfectly rational and predictable. The authors built a "Relay Table" (like a subway map).
  • Start: No one believes anything.
  • Step 1: If Rumor 1 is strong enough, it takes over.
  • Step 2: If Rumor 2 is also strong enough, it joins the party.
  • Step 3: If people get tired and quit, the system shifts again.
• The Magic: They proved that for every single step in this chain, the math is exact. You can calculate exactly when the baton passes from one state to the next.

Scenario B: The Complex Case ($\omega > 0$)

• The Setup: Now, skeptics can be "re-activated" or influence others to quit. This adds a feedback loop (like a snake eating its own tail).
• The Result: The math gets messy. You can't write down a simple formula for the new states anymore (they involve "irrational" numbers, like square roots).
• The Triumph: Even though the formulas are messy, the Relay Structure remains! The authors showed that even in this complex, messy world, the system still follows the same "baton pass" rules. The instability of the old state still guarantees the existence of the new state. They used a "rank-one perturbation" (a fancy way of saying "adding one small feedback loop") to prove the system stays stable enough to follow the relay path.

Why This Matters (The "So What?")

1. It Unifies Fields: It connects Chemistry, Epidemiology, and Ecology. It shows that the way molecules react, viruses spread, and rumors travel all follow the same "Relay" logic.
2. It Predicts the Future: Instead of running thousands of computer simulations to guess what happens, you can look at the "Siphon Lattice" (the map of walls) and calculate exactly which states are possible and which are impossible.
3. It Prevents Surprises: It tells us that in these types of systems, you won't get chaotic, unpredictable jumps. The system moves in a logical, step-by-step progression. If a rumor dies, it's because a specific threshold was crossed, and a specific new state is waiting to take over.

Summary Analogy

Imagine a domino tower built on a grid.

• Old View: You push one domino, and it falls. You don't know which one hits next.
• This Paper's View: The dominoes are arranged in a specific "Siphon Lattice." When the first domino falls (becomes unstable), it only hits the specific domino directly next to it in the lattice. The force required to knock the first one down is exactly the force needed to set up the second one.
• The Tool: The authors built a "Relay Map" that shows you the entire path of the falling dominoes before you even push the first one.

This paper gives us a "Relay Map" for complex social and biological systems, turning a chaotic mess of equations into a clear, predictable path of evolution.

Technical Summary

Here is a detailed technical summary of the paper "Relay transitions and invasion thresholds in multi-strain rumor models: a chemical reaction network approach" by Florin Avram and Andrei-Dan Halanay.

1. Problem Statement

The paper addresses the complexity of analyzing positive Ordinary Differential Equation (ODE) models arising in mathematical epidemiology (ME), ecology, and rumor spreading on Online Social Networks (OSN). Specifically, it tackles the challenge of understanding multi-strain dynamics where multiple competing "strains" (e.g., different rumors or pathogens) interact.

Traditional analysis often focuses on the Disease-Free Equilibrium (DFE) and uses the Next-Generation Matrix (NGM) to determine a single basic reproduction number ($R_0$). However, in multi-strain systems:

• Equilibria exist on various boundary faces (where some strains are extinct).
• Stability transitions between these equilibria are complex and often involve transcritical bifurcations.
• There is a lack of a unified framework connecting the combinatorial structure of the network (Chemical Reaction Networks - CRN) with the dynamical stability analysis (ME).

The authors aim to formalize the mechanism by which a resident equilibrium loses stability and a "successor" equilibrium (with an additional strain) emerges, a phenomenon they term a "boundary transcritical relay."

2. Methodology

The authors employ a hybrid methodology combining Chemical Reaction Network Theory (CRNT), Mathematical Epidemiology, and Symbolic Computation.

• CRN Framework: They model the OSN rumor dynamics as a Chemical Reaction Network. Key to this approach is the identification of Siphons (or semi-locking sets). A siphon is a set of species such that if they are absent, they cannot be produced. In ODE terms, the face where these species are zero is forward-invariant.
• Siphon Lattice: The authors organize the system's equilibria not as isolated points, but as nodes within a lattice generated by minimal siphons. Moving up or down this lattice corresponds to adding or removing entire blocks of species (strains).
• Transversal Jacobian Blocks: By exploiting the forward invariance of siphon faces, the Jacobian matrix at an equilibrium can be decomposed into block structures. The stability of an equilibrium against the invasion of a specific siphon block is determined by the transversal Jacobian block ($M_\sigma$).
• Reproduction Functions & Invasion Numbers: They generalize the concept of $R_0$ to reproduction functions defined on siphon faces. Evaluating these functions at specific resident equilibria yields invasion numbers.
• Symbolic Package (EpidCRN): The authors utilize their symbolic software package, EpidCRN, to automate the decomposition of ODEs into CRNs, compute minimal siphons, derive reproduction functions, and perform exact symbolic stability analysis (using Routh-Hurwitz criteria) to avoid numerical errors.
• Case Study: The methodology is applied to a specific 8-species, 13-parameter OSN rumor model (Fakih et al., 2025) involving two competing rumors, user skepticism, and platform withdrawal. They analyze two cases:
  1. $\omega = 0$: No feedback from skeptics to withdrawn users (rational equilibria).
  2. $\omega > 0$: Feedback exists (irrational equilibria requiring perturbation analysis).

3. Key Contributions

A. The "Relay" Mechanism

The paper introduces the concept of a Boundary Transcritical Relay (BTR). They prove that for "ME-type" siphons (where the transversal Jacobian is a Metzler matrix admitting a regular splitting):

• Coincidence of Inequalities: The condition for a resident equilibrium to lose transversal stability (invasion number $> 1$) is identical to the condition for the existence of a successor equilibrium on the adjacent face (where the invading species are positive).
• Structural Origin: This is not a coincidence but a structural consequence of:
  1. Face invariance induced by siphons.
  2. Factorization of the vector field in transversal variables.
  3. Metzler structure of the Jacobian blocks.

B. Siphon-Lattice Relay Graphs

The authors propose visualizing global boundary dynamics as a directed relay graph:

• Nodes: Equilibria indexed by inhabited siphon faces.
• Edges: "Distance-one" moves in the siphon lattice (releasing one minimal siphon block).
• Labels: Explicit invasion numbers derived from reproduction functions.
  This provides an algorithmic way to enumerate equilibria and predict stability transitions as parameters vary.

C. Algorithmic Detection

They provide an exact algorithm (implemented in Mathematica) to detect relay transitions:

1. Compute equilibria on a face.
2. Calculate the spectral abscissa of the transversal block.
3. If the abscissa crosses zero, verify the existence and stability of the successor on the adjacent face.
   This algorithm succeeds if all intermediate expressions are rational; otherwise, it flags "Undecided."

D. Generalization of NGM Theorem

The paper generalizes the classic Next-Generation Matrix theorem. It proves that for any forward-invariant face defined by a siphon, the mixed Jacobian block vanishes, forcing a block lower-triangular structure. This allows stability to be reduced to the spectral properties of the transversal block, linking CRN structure directly to stability thresholds.

4. Key Results

The OSN Model Analysis ($\omega = 0$)

• Rational Equilibria: All 9 equilibria (including DFE, rumor-free, and single/dual-strain endemic states) admit explicit rational formulas.
• Relay Table: The authors construct a complete "relay table" showing the exact parameter thresholds where one equilibrium destabilizes and another emerges.
• Impossibility of Oscillations: They prove that for $\omega = 0$, the system cannot undergo Hopf bifurcations. The Jacobian has a block lower-triangular structure where the "platform" dynamics (users) are autonomous and the "strain" dynamics are driven slaves. This structural decoupling prevents the complex conjugate eigenvalues required for oscillations.

The OSN Model Analysis ($\omega > 0$)

• Irrational Equilibria: When feedback ($\omega > 0$) is introduced, equilibria become irrational (roots of quadratic equations).
• Persistence of Relay Structure: Despite the loss of explicit formulas, the relay structure remains valid. The invasion inequalities governing the transitions are independent of $\omega$ because the strain-block equations do not contain $\omega$.
• Stability Verification: Using rank-one perturbation analysis (Matrix Determinant Lemma), they derive sufficient conditions for the stability of the new equilibria, showing that the relay framework predicts the system's behavior even when exact coordinates are unavailable.

Comparison with Classical Bifurcations

The paper distinguishes Transcritical Relays from classical transcritical bifurcations in four ways:

1. Geometric Origin: The critical eigenvalue is enforced by the invariant face geometry, not arbitrary parameter unfolding.
2. Block Invasion: Invasion occurs via a block of species (a siphon), not a single direction.
3. Lattice-Prescribed Successor: The successor is constrained to a specific adjacent face in the siphon lattice.
4. Shared Inequality: A single inequality governs both the instability of the resident and the existence of the successor.

5. Significance

• Unification of Disciplines: The paper successfully unifies CRNT (siphons), ME (reproduction numbers), and Ecology (invasion graphs) into a single coherent framework for analyzing positive ODEs.
• Algorithmic Predictability: It moves the analysis of multi-strain models from case-by-case algebraic manipulation to a systematic, algorithmic procedure based on the siphon lattice.
• Handling Complexity: It demonstrates how to analyze complex, non-rational systems (like the $\omega > 0$ case) by relying on structural properties (invariant faces and Metzler blocks) rather than explicit solutions.
• Practical Application: The findings provide a rigorous method for predicting rumor spreading outcomes, identifying critical thresholds for rumor extinction or dominance, and understanding how platform features (like skepticism or withdrawal) alter the stability landscape.
• Software Development: The work validates and showcases the utility of the EpidCRN symbolic package for automating complex stability analyses in biological networks.

In conclusion, the paper establishes that the "coincidence" between instability and existence in multi-strain models is a fundamental structural feature of positive dynamical systems with invariant faces, governed by the combinatorics of siphons. This insight allows for the construction of "relay graphs" that map the global dynamics of complex interaction networks.