Spontaneous fine-tuning to environment in many-species chemical reaction networks

Abstract

A chemical mixture that continually absorbs work from its environment may exhibit steady-state chemical concentrations that deviate from their equilibrium values. Such behavior is particularly interesting in a scenario where the environmental work sources are relatively difficult to access, so that only the proper orchestration of many distinct catalytic actors can power the dissipative flux required to maintain a stable, far-from-equilibrium steady state. In this article, we study the dynamics of an in silico chemical network with random connectivity in an environment that makes strong thermodynamic forcing available only to rare combinations of chemical concentrations. We find that the long-time dynamics of such systems are biased toward states that exhibit a fine-tuned extremization of environmental forcing.A n arrangement of matter may be said to be finely tuned to its environment if it is configured to interact with that environment in a way that is highly atypical for random rearrangements of its components. Whereas some of the most striking examples of such matching between system and environment are found in the architecture of living things (1, 2), the category is in principle far broader, and one might therefore hope to provide an account of emergent fine-tuning in general physical terms.Recent progress in theoretical nonequilibrium statistical mechanics (3-7) has helped to clarify a general relationship between the likelihood of a system adopting a particular microscopic configuration and the amount of energy absorbed and dissipated during the system's dynamical history. Essentially, the irreversibility of being driven into a kinetically stable state must be compensated by the release of heat in the surroundings, which frequently occurs when the system absorbs and dissipates energy from an environmental source of work. The implications of this thermodynamic frame are particularly clear in systems where the specific configuration adopted has a large influence on the accessibility of energy from external drives, such that strongly driven states are rare. In such a scenario, having a highly dissipative history becomes tantamount to having a configurational history that includes the rare, finely tuned states that make dissipation possible. This observation suggests that there may be a broad class of nonequilibrium systems for which state-dependent drive strength is the relevant thermodynamic parameter for describing the bias in their exploration of configuration space.Motivated by these considerations, we carried out simulations of a complexly driven, randomly wired, many-species chemical reaction network, with the aim of identifying a thermodynamic principle governing self-organization far from equilibrium. We found that such networks have a pronounced bias toward settling into rare states in chemical space with extremal forcing, suggesting that emergent fine-tuning is a general tendency of their dynamical behavior. Two-Species ModelIn this study, we are ultimately interested in the thermodynamic propert...

Onset of natural selection in auto-catalytic heteropolymers

“…In the Kauffman case, some of the molecules types in the pool are capable of catalyzing the synthesis of others from "raw materials" (abundant small metabolites) ultimately resulting in the emergence of metabolic autocatalytic cycles in large systems. A recent model of such chemical reactions [30] shows that the system self-organizes to a state finely tuned to the external driving force. This can be interpreted as maximization of the rate of negentropy adsorption from the environment.…”

Becoming Cognitive Science

“…Triangulating on the nature of minds by considering the commonalities shared by their biological, mechanical, and community forms provides constraints that a single perspective cannot provide on its own. By striving to create a general enough conception of minds to capture their diverse manifestations, cognitive science belongs to the relatively uncommon class of “integrative sciences.” This class also includes artificial life—creating software, robotic, and synthetic biological life forms to understand the essential nature of life abstracted beyond its DNA‐based forms (Agmon, Gates, Churavy, & Beer, ; Bedau, ; Horowitz & England, ). A third exemplar is complex systems, the general study of systems made up of parts that interact to create emergent, global patterns such as positive and negative feedback loops, oscillations, and recursive branching structures (Mitchell, ; Page & Miller, ; Thurner, Klimek, & Hanel, ).…”

Infinitely large, randomly wired sensors cannot predict their input unless they are close to deterministic

“…sparsity in random reaction networks was key to the results of Ref. [24]. Our results suggest that given a particular inherent sensor stochasticity, there is an optimal finite sensor size at which functionality (prediction) is maximized.…”