A generic neutral model for measuring excess evolutionary activity of genotypes.pdf

In W. Banzhaf, J. Daida, A. E. Eiben, M. H. Garzon, V. Honavar, M. Jakiela, R. E. Smith, eds., GECCO’99: Proceedings of the Genetic and Evolutionary Computation Conference, Vol. 2 (pp. 1366-1373), San Francisco: Morgan Kaufmann, 1999. A Generic Neutral Model for MeasuringExcess Evolutionary Activity of GenotypesAndreas RechtsteinerSystems Science Ph.D. ProgramPortland State UniversityP.O. Box 751, Portland OR 97207andreas@sysc.pdx.edu(503) 725-4992 Mark A. BedauDepartment of PhilosophyReed College3203 SE Woodstock Blvd, Portland OR 97202mab@reed.edu(503) 771-1112, ext. 7337AbstractWe introduce and study a simple genericmodel of neutral evolution of genotypes, de-signed to provide a feasible and generalmethod for quantifying excess evolutionaryactivity|the extent to which evolutionaryactivity is the product of adaptive evolu-tion. We compare the behavior of the genericneutral model against two other models:Packards agent-based model of the evolutionof sensory-motor functionality and a neutral\shadow of Packards model. Diversity andevolutionary activity of these three modelsacross the mutation rate spectrum illustratethe feasibility and general applicability of thegeneric neutral model, conrm the appropri-ateness of using neutral models to quantifythe extent of the continual adaptive success ofgenotypes, and reveal power-law dependencesof evolutionary activity on mutation rate.1 The Need for Neutral ModelsAlthough it is commonly accepted that adaptive evo-lution produces much of the structure and function-ality in complex systems [6, 4], it is often dicult todistinguish adaptive change from other evolutionaryphenomena such as random genetic drift [3]. Someeven question whether adaptations can be objectivelyidentied at all [3]. The ultimate goal of this paperis to facilitate the investigation of universal laws ofadaptive evolution. Toward this end, this paper aimsto develop statistics for objectively identifying andquantifying adaptive evolutionary activity, especi