ABSTRACT
This chapter introduces a simple model of molecular evolution based upon spin-glass concepts and discusses its behavior. It identifies adaptive and neutral evolutionary behaviors. They are separated by a transition, which bears some relation with the gelation transition in spin glasses. A massive example of parallel computation goes under our own eyes, but is too slow to detect: evolution. M. Eigen et al. have considered deterministic equations governing the evolution of the genotype distribution in a population of self-replicating molecules. W. Fontana and P. Schuster have introduced a computer model of population evolution in a given fitness landscape, which models the dynamics of replication of ribonucleic acid-like molecules in prebiotic conditions. They are able to highlight several interesting features of the evolution mechanism, but their conclusions are restricted by the choice of a particular fitness function. S. A. Kauffman and S. Levin have introduced a class of fitness functions with tunable correlations in an evolutionary context.
