Richard Bagley

Paper #: 91-04-020

We present a continuation of the investigation of autocatalytic chemical reaction networks as proposed by Kauffman and begun by Farmer, Kauffman, and Packard. Autocatalytic reaction networks are defined as a set of reversible ligation reactions between polymers. All reactions are catalyzed by at least one polymer which also participates as a reactant in the network. The model was motivated by the problem of the origin of life, and in particular, the transition from no-life (abiogenesis) to life (biogenesis). This transition is modeled as the self-organization of polymers from a world of spontaneously generated momers. We have made several significant extensions to the model, including a treatment of the uncatalyzed processes which compete with the network for resources, a consistent and physically meaninfgul chemistry, and a source of variability which has conferred the capacity for open ended evolution. We explore the properties of these networks, and discuss the requirements for successful production of large polymers and stability to variations in the environment. We also present results on constraints on parameters which represent the environment or dynamic properties of the network. By attending to the need to explain such problems, we have a model of functional self-organization which is motivated by a real problem, albeit a poorly understood one. We also present results on a number of related issues, such as the theory of evolution, the emergence of collective rules of behavior from local rules of interaction, network dynamics and connectivity, and the relevance of our studies to real experiment.