Understanding the relationship between genotype and phenotype is essential for understanding evolutionary dynamics. RNA is an extremely simple yet highly instructive case: both genotype (sequence) and phenotype (structure) are different aspects of a single molecule. The influence of the generic RNA folding properties on the dynamics of selection deserves attention, given the successful in vitro evolution of RNA sequences to optimize replication rates [1, 2], or to perform recognition [3, 4] as well as catalytic tasks [5, 6, 7, 8]. We study a computer model of a population of replicating and mutating RNA sequences which are subject to selection of their secondary structure. Dominant phenotype and fitness undergo sudden transitions between long periods of stasis. In contrast, at the genetic level the population can be described by a diffusion process in sequence space along extended networks of structurally equivalent sequences. The population has a complex fine structure consisting of numerous well-separated clusters which move independently and undergo random selection. The present work provides a mechanistic model for the assumptions underlying and the phenomenology resulting from the Neutral Theory of molecular evolution [9, 10].
Return to 1995 working papers list.