Reaction Systems and Computational Dynamics

Reaction systems constitute a formal paradigm inspired by the combinatorial interplay of biochemical reactions in living cells. At their core, a reaction system is defined by a finite set of entities and a collection of reaction rules, each specified by reactants, inhibitors and products. The dynamics unfolds in discrete steps: at each update, enabled reactions—those whose reactants are present and inhibitors absent—execute simultaneously to yield a new set of entities. This simplistic yet expressive framework supports the rigorous analysis of system behaviour, including the identification of fixed points and attractors, and it can be extended to richer computational devices known as reaction automata that accept formal languages. Recent advances have explored semantic variants—such as asynchronous or maximally permissive updates—that bring reaction systems into correspondence with Boolean networks and other models of regulatory biology. Computational studies have characterised decidability and complexity in subclasses of reaction systems, revealing polynomial-time solvable cases alongside intractable settings. On the applied side, reaction systems have been used to model pivotal biological pathways, from haemostatic cascades to immune differentiation, and to design programmable biochemical circuits that mimic memory and learning. This duality—between abstraction and real-world application—underscores the global significance of reaction systems as a bridge between theoretical computer science and molecular biology.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Research from all publishers

Studies on additive reaction systems have demonstrated that when each reaction involves at most one reactant and no inhibitors, key decision problems such as the existence of fixed points and periodic attractors become polynomially solvable. By reducing reaction dynamics to graph-theoretic representations, researchers have identified efficient algorithms that may inform the analysis of more general systems. In parallel, work on pure reaction automata has introduced a maximal-reactive update scheme in which unconsumed entities are discarded at each step; this has led to new hierarchies of computational power, showing that non-deterministic variants achieve Turing completeness while deterministic ones do not. Finally, the adoption of reaction systems to model the secondary haemostasis pathway has yielded a discrete framework for blood-clotting dynamics. By encoding key enzymes and inhibitors as reaction rules, investigators have used computational tools to simulate thrombus formation and to predict the impact of molecular perturbations, thereby offering a platform for in silico exploration of therapeutic interventions.

Reaction Systems and Computational Dynamics publication trend

The graph below shows the total number of articles in reaction systems and computational dynamics across all publications each year (not limited to Nature Index journals).

Technical terms

Reaction system: A formal model comprising entities and reaction rules that synchronously updates the set of entities based on reactants, inhibitors and products.

Entity: An abstract symbol or molecular species present in a system state, serving as reactant, inhibitor or product.

Inhibitor: An entity whose presence prevents a reaction from occurring.

Facilitator: An entity whose presence is required to enable a reaction.

Fixed point: A state that remains unchanged under the update rules of the reaction system.

Attractor: A state or cycle of states toward which the system inevitably evolves over time.

References

1. DNA Reaction System That Acquires Classical Conditioning. ACS Synthetic Biology (2024).
2. Fixed points and attractors of additive reaction systems. Natural Computing (2024).
3. Chemical pure reaction automata in maximally parallel manner. Journal of Membrane Computing (2025).
4. A Computational Model of the Secondary Hemostasis Pathway in Reaction Systems. Mathematics (2024).
5. Melding Boolean networks and reaction systems under synchronous, asynchronous and most permissive semantics. Natural Computing (2024).