Structure and Input-to-State Stability for Composable Computations in Chemical Reaction Networks

Basic Information

• Paper ID: 2510.13116
• Title: Structure and input-to-state stability for composable computations in chemical reaction networks
• Authors: Renlei Jiang (Zhejiang University), Chuanhou Gao (Zhejiang University), Denis Dochain (KU Leuven, Belgium)
• Classification: math.DS (Dynamical Systems)
• Publication Date: October 15, 2025 (arXiv preprint)
• Paper Link: https://arxiv.org/abs/2510.13116

Abstract

In the field of molecular computation based on chemical reaction networks (CRNs), maintaining predefined computational functionality in coupled mass-action systems (MASs) through parallelism has been a research focus. MASs possessing this property are termed composable. This paper investigates structural conditions for the composability of two MASs. By leveraging input-to-state stability (ISS) properties, we identify a specific class of CRN architectures that guarantee composability with other networks. A concrete example validates the conclusions and illustrates the application of composability in computing composite functions.

Research Background and Motivation

Core Problem

The fundamental problem addressed in this research is: How can we determine structural conditions of chemical reaction networks such that multiple mass-action systems can execute molecular computations in a composable manner?

Problem Significance

1. Parallelism Challenge in Molecular Computation: Chemical reactions are inherently parallel, while traditional computation requires sequential execution. This fundamental contradiction is the core challenge facing molecular computation.
2. Practical Value: Molecular computation has tremendous application potential in biocompatible environments such as medical diagnostics and data storage.
3. Theoretical Completeness: While MASs have been proven to be Turing-complete, systematic theory for implementing modular composition of complex computations remains lacking.

Limitations of Existing Approaches

1. Chemical Oscillator Methods: Although capable of sequential execution, networks are overly complex and introduce significant errors in multi-step computations.
2. Rate-Independent CRNs: Computational capabilities are severely limited, unable to meet complex computation requirements.
3. Lack of Structural Analysis: Existing research primarily relies on dynamical analysis, lacking methods to directly determine composability from network structure.

Research Motivation

Building on the ISS criterion framework established by Jiang et al. (2025), this paper aims to establish direct connections between network structure and composability, providing more easily verifiable structural conditions.

Core Contributions

1. Established Structure-Composability Relationship: First directly linking structural properties of chemical reaction networks with composability.
2. Proposed Verifiable Sufficient Conditions: Structural conditions based on weak reversibility, zero deficiency, and mass conservation (Theorem 10).
3. Constructed ISS-Lyapunov Function Theory: Employing pseudo-Helmholtz free energy functions as ISS-Lyapunov functions.
4. Provided Concrete Application Examples: Demonstrated how to utilize composability for computing composite functions.

Methodology Details

Task Definition

A mass-action chemical reaction computer (msCRC) is defined as a six-tuple:

C = (S, C, R, κ, X, Y)

where:

• S: set of species
• C: set of complexes
• R: set of reactions
• κ: reaction rate constants
• X ⊂ S: set of input species
• Y = S \ X: set of output species

Dynamic Computation: For a positive function σ : ℝᵐ₍≥₀₎ → ℝⁿ⁻ᵐ₍≥₀₎, an msCRC executes dynamic computation of σ if and only if:

lim_{t→∞} y(t) = σ(x₀)

Core Theoretical Framework

1. Composability Definition

Two msCRCs C₁ and C₂ are dynamically composable if they satisfy:

• (A.1) Y₁ = X₂, Y₂ ∩ X₁ = ∅ (output-input matching)
• (A.2) Each possesses specific dynamical forms
• (A.3) Each can reach desired steady states

2. ISS Criterion (Theorem 5)

If the y₂-related portion of C₂ satisfies the ISS condition with respect to (ȳ₁, ȳ₂):

|y₂(t) - ȳ₂| ≤ β(|y₂₀ - ȳ₂|, t) + γ(sup_{0≤τ≤t} |x₂(τ) - ȳ₁|)

then C₁ and C₂ are dynamically composable.

Main Technical Innovations

1. Reduced System Concept

For an msCRC C = (S, C, R, κ, X, Y), its reduced system C̃ = (S̃, C̃, R̃, κ̃(t)) is defined as:

• S̃ = Y (retaining only output species)
• C̃ = π_Y(C) (projection of complexes onto output species)
• Reaction rates κ̃_j(t) = κ_j ∏{S_i∉S̃} s_i^{v}(t) (input species dynamics embedded in rate constants)

2. Structural Conditions Theorem (Theorem 10)

Main Result: If the reduced system C̃₂ satisfies:

1. Weak Reversibility: Each connected component is strongly connected
2. Single Connected Component with Zero Deficiency: δ = n - l - dim S = 0
3. Mass Conservation: ∃v ∈ ℝⁿ₍>₀₎ such that v^T g₂(x₂, y₂) = 0

then C₁ and C₂ are dynamically composable.

3. ISS-Lyapunov Function Construction

Utilizing the pseudo-Helmholtz free energy function:

V(s, s̄) = Σⱼ₌₁ʳ (sⱼ(ln sⱼ - ln s̄ⱼ - 1) - s̄ⱼ)

as an ISS-Lyapunov function, combined with the zero deficiency theorem to guarantee stability.

Experimental Setup

Example System Design

The paper validates theoretical results through a concrete molecular computation system:

System 1 (Summation Computation):

X₁ →¹ X₁ + Y₁, X₂ →¹ X₂ + Y₁, Y₁ →¹ ∅
X₃ →¹ X₃ + Y₂, X₄ →¹ X₄ + Y₂, Y₂ →¹ ∅

Computed function: σ₁(x₁,x₂,x₃,x₄) = (x₁+x₂, x₃+x₄)^T

System 2 (Normalization Computation):

Y₁ + Z₁ →¹ Y₁ + Z₂, Y₂ + Z₂ →¹ Y₂ + Z₁

Computed function: σ₂(y₁,y₂) = (y₁/(y₁+y₂), y₂/(y₁+y₂))^T

Verification Methods

1. Structural Condition Verification: Checking whether the reduced system satisfies the three conditions of Theorem 10
2. Numerical Simulation: Verifying whether the composite system correctly computes the composite function
3. Convergence Analysis: Confirming that the system reaches the expected steady state

Experimental Results

Main Results

1. Structural Condition Verification

The reduced system of System 2:

Z₁ ⇌^{y₁(t)}_{y₂(t)} Z₂

satisfies all structural conditions:

• ✓ Weak reversibility (strongly connected)
• ✓ Single connected component with zero deficiency (δ = 2-1-1 = 0)
• ✓ Mass conservation (v = (1,1) such that v^T g₂ = 0)

2. Composite Function Computation

The composite system successfully computes the composite function:

σ(x₁,x₂,x₃,x₄) = ((x₁+x₂)/Σxᵢ, (x₃+x₄)/Σxᵢ)

3. Numerical Simulation Results

Using initial values x₁(0)=0.2, x₂(0)=0.3, x₃(0)=0.6, x₄(0)=0.1:

• Theoretical expectation: (0.5/1.2, 0.7/1.2) ≈ (0.417, 0.583)
• Simulation result: System converges to expected values, validating theoretical correctness

Key Findings

1. Structure Determines Composability: Network topology directly determines composability without requiring analysis of complex dynamics.
2. Advantages of Zero Deficiency Networks: Zero deficiency conditions guarantee the existence of ISS-Lyapunov functions.
3. Importance of Mass Conservation: Mass conservation ensures system boundedness and stability.

Related Work

Main Research Directions

1. Rate-Independent CRNs: Chalk et al. (2019) formalized the composability concept, but with limited computational capabilities.
2. Stochastic CRNs: Severson et al. (2019) extended to stochastic cases.
3. MAS Composability: Jiang et al. (2025) ISS criterion-based approach, upon which this paper builds.

Advantages of This Work

1. Direct Structural Conditions: Structural conditions are more easily verifiable compared to dynamical analysis.
2. Broader Applicability: MASs possess Turing completeness with stronger computational capabilities.
3. Theoretical Completeness: Provides rigorous mathematical proofs of sufficient conditions.

Conclusions and Discussion

Main Conclusions

1. Established Structure-Function Relationship: Network structures with zero deficiency, weak reversibility, and mass conservation guarantee composability.
2. Simplified Verification Process: Direct determination of composability from network topology, avoiding complex dynamical analysis.
3. Guidance for Molecular Computation Design: Can be used to construct "composable basic msCRC libraries."

Limitations

1. Sufficient but Not Necessary Conditions: Theorem 10 provides sufficient conditions, which may be overly conservative.
2. Single Architecture Type: Only considers zero deficiency networks; other structural types remain unexplored.
3. Practical Implementation Challenges: Gap remains between theoretical results and physical implementation in DNA strand displacement reactions.

Future Directions

1. Extend Network Architectures: Explore broader types of composable network structures.
2. Build Computation Libraries: Establish composable basic msCRC libraries.
3. Optimize Design: Research how to design more efficient composable molecular computation systems.

In-Depth Evaluation

Strengths

1. Significant Theoretical Contribution: First to establish direct connection between network structure and composability, with important theoretical value.
2. Strong Method Innovation: Cleverly combines zero deficiency theorem with ISS theory.
3. Sufficient Verification: Rigorous theoretical derivation with clear example validation.
4. High Practical Value: Provides actionable guidance for molecular computation system design.

Weaknesses

1. Limited Applicability: Only applicable to networks satisfying specific structural conditions.
2. Lack of Necessity Analysis: Does not discuss necessity of proposed conditions.
3. Large-Scale Network Handling: Practical application to large-scale networks requires further research.

Impact

1. Academic Value: Provides important structured analysis tools for molecular computation theory.
2. Application Prospects: Can guide design and implementation of practical molecular computation systems.
3. Reproducibility: Clear theoretical results, easily verifiable and extensible.

Applicable Scenarios

1. Molecular Computation System Design: Provides theoretical foundation for designing composable molecular computation modules.
2. Biochemical Network Analysis: Can analyze modularity characteristics of natural biochemical networks.
3. Synthetic Biology: Guides design of functional modules in synthetic biology.

References

The paper cites important literature in the field, including:

• Feinberg (1987): Foundational work on zero deficiency theorem
• Chalk et al. (2019): Formalization of composability concept
• Jiang et al. (2025): Prior work on ISS-based composability analysis
• Chaves (2005): Application of ISS-Lyapunov functions in biochemical networks

Overall Assessment: This is a high-quality theoretical paper making important contributions to the field of molecular computation. By combining network structure theory with control theory, it provides new analytical tools for composability in molecular computation systems. While having certain limitations in applicability, it establishes important theoretical foundations for further development in this field.