Thermodynamic patterns of life: emergent phenomena in reaction networks

Reaction networks are an important tool for the analysis of complex chemical reaction systems. They help us understand systems ranging from specific metabolisms to planetary atmospheres. This thesis develops methods for the analysis of living systems by using reaction networks with a focus on the inclusion of thermodynamic properties. New methods for more realistic artificial chemistries are developed using thermodynamic constraints. A model of evolvable artificial ecosystems is created to understand the effect of evolution and life on the flow of matter and energy through the system. To investigate general thermodynamic properties of large-scale reaction networks, artificial reaction networks are created with a simple scheme for deriving thermodynamically consistent reaction rates. Linear and nonlinear networks using four different complex network models are simulated to their non-equilibrium steady state for various boundary fluxes. Increasing the flow through nonlinear networks shows to increases the number of cycles and leads to a narrower distribution of chemical potentials. In the context of finding signs of life by detecting chemical disequilibrium, a photochemical model of the modern atmosphere and a model of the Archean atmosphere are compared. Calculating the reaction pathways that are most relevant for explaining their reaction network's steady state with a new method allows for the detection of topological differences between the two models. Pathways of the modern Earth atmosphere are simpler (less reactions) and contain fewer cycles than their Archean counterparts. To model the influence of life on reaction pathways, an artificial ecosystem model is developed. Evolution of the reaction networks entails an evolution of reaction pathways towards simplicity, thus indicating that the presence of pronounced, relatively simple pathways in real systems is a consequence of an evolutionary mechanism