The SHAPE of logistic growth shows that serial passaging biases fixation probability

Abstract

The forward time simulation tool rSHAPE (R-package for Simulated Haploid Asexual Population Evolution) was designed to complement the theoretical and empirical study of evolution. Included with rSHAPE are functions to programmatically build, run, and initially process results of an evolutionary experiment defined by the range of experimental conditions. As experimental evolution often studies both the emergence and fate of de novo mutants, I validated rSHAPE by confirming its ability to replicate seminal theoretical expectations concerning changes in fitness through time and the fixation probability of mutants. As an example of how rSHAPE can support both theoretical and empirical research, I applied rSHAPE to study how the laboratory protocol of serial passaging, common in microbial experimental evolution, affects the fixation probability of de novo mutants. Unlike related theoretical work which modelled growth as effectively exponential (Wahl et al., 2002), this study considered populations experiencing logistic growth which is common to microbes undergoing serial passaging. In contrast to exponential growth, when a population undergoes logistic growth the probability of a mutant arising and eventually fixing depends upon when a mutant arises during a growth phase. Users can download software and documentation for rSHAPE through CRAN at https://cran.r-project.org/web/packages/rSHAPE/index.html, or via GitHub at https://www.github.com/Jdench/SHAPE_library/.