Uniqueness of feasible equilibria for mass action law (MAL) kinetic systems

This paper studies the relations among system parameters, uniqueness, and stability of equilibria, for kinetic systems given in the form of polynomial ODEs. Such models are commonly used to describe the dynamics of nonnegative systems, with a wide range of application fields such as chemistry, systems biology, process modeling or even transportation systems. Using a flux-based description of kinetic models, a canonical representation of the set of all possible feasible equilibria is developed. The characterization is made in terms of strictly stable compartmental matrices to define the so-called family of solutions. Feasibility is imposed by a set of constraints, which are linear on a log-transformed space of complexes, and relate to the kernel of a matrix, the columns of which span the stoichiometric subspace. One particularly interesting representation of these constraints can be expressed in terms of a class of monotonous decreasing functions. This allows connections to be established with classical results in CRNT that relate to the existence and uniqueness of equilibria along positive stoichiometric compatibility classes. In particular, monotonicity can be employed to identify regions in the set of possible reaction rate coefficients leading to complex balancing, and to conclude uniqueness of equilibria for a class of positive deficiency networks. The latter result might support constructing an alternative proof of the well-known deficiency one theorem. The developed notions and results are illustrated through examples.

💡 Research Summary

This paper, titled “Uniqueness of feasible equilibria for mass action law (MAL) kinetic systems,” presents a comprehensive analysis of the relationships between system parameters, uniqueness, and stability of equilibrium points in kinetic systems modeled by polynomial ordinary differential equations (ODEs). These models, central to Chemical Reaction Network Theory (CRNT), describe nonnegative systems with wide applications in chemistry, systems biology, process engineering, and beyond.

The core contribution of the work is the development of a canonical, flux-based representation for the set of all feasible equilibrium states in weakly reversible mass action systems. The authors reformulate the system dynamics using a structure where the time evolution of species concentrations is expressed as the product of a matrix whose columns span the stoichiometric subspace and a vector function linked to concentrations through a class of strictly stable compartmental (Metzler) matrices. This representation provides a novel framework for equilibrium analysis.

A key insight is the characterization of feasibility—the condition that an equilibrium lies within the positive orthant of concentration space. The authors show that feasibility can be imposed by a set of constraints that become linear in a log-transformed space of reaction complexes. These constraints relate to the orthogonality between a log-transformed vector function and the kernel of the stoichiometric matrix. Notably, these constraints can be equivalently expressed in terms of a class of monotonically decreasing functions.

This monotonicity property serves as a powerful tool to establish connections with classical results in CRNT. First, it allows for the identification of regions in reaction rate parameter space that lead to complex-balanced equilibria, a well-studied and stable class of steady states. Second, and more significantly, the monotonicity is employed to conclude the uniqueness of equilibria for a specific class of reaction networks with positive deficiency. This result provides a potential pathway towards constructing an alternative proof of the celebrated Deficiency One Theorem, a cornerstone of CRNT that gives sufficient structural conditions for equilibrium uniqueness in each positive stoichiometric compatibility class.

The paper is structured to first introduce the formal description of reaction network structure and dynamics. It then details the development of the flux-based canonical representation and the derivation of the feasibility conditions. Subsequent sections explore the link between network structure and the monotonicity of these feasibility functions, culminating in a discussion that bridges these new insights with established CRNT literature on existence, uniqueness, and stability.

By combining graph-theoretic concepts (like linkage classes and weak reversibility) with algebraic and analytic methods, the research offers a unified perspective on equilibrium properties. The proposed framework not only advances theoretical understanding but also holds practical promise for applications such as analyzing multistability in biological networks or designing stable operating regimes in open process systems through control and optimization. The arguments are supported and illustrated with examples throughout the text.