Parameters Estimation for Stochastic Genetic Evolution of Asexual Populations



Journal Title

Journal ISSN

Volume Title



The evolution of asexual microorganisms depends primarily on their selective advantages (fitnesses) and mutation rates. Previous attempts to estimate these parameters failed for realistic cases with multiple mutations. This work proposes a new method of extracting key parameter features. The prime goal is to develop and verify this method. The objectives are: to make a stochastic model; to extract parameters via fitting frequency curves; to find estimates for the parameters; to expand the methodology to multiple events; to evaluate the estimation quality.

The fitting errors were of the order of 10^(-2). Depending on the number of runs per simulated experiment, ranging between 10^2 and 10^5, the successful prediction of the number of possible mutants was within 78% and 99.4%, respectively. Their selective advantage estimates had relative errors of 2%-8%. The estimation algorithm was consistent as it withstood different parameter alterations in experiment sizes, mutation rates, and transition probability distributions. The algorithm was verified for up to 6 possible genotypes.

The now-verified method can be applied to the understanding of growth patterns of microorganisms (such as Escherichia coli). Further, it can be generalized so as be applied to more complex species.



Stochastic, Random processes, Rare events, Modeling, Parameter estimation, Algorithms, Genetics, Adaptive, Evolution, Biomath, Mathematics, Mutations, Mutant, Asexual populations, E. coli, Antigens, Bacteria