Study of selected phenotype switching strategies in time varying environment

• Published in: Physics letters. A 2016-03, Vol.380 (13), p.1267-1278
• Main Authors: Horvath, Denis, Brutovsky, Branislav
• Format: Article
• Language: English
• Subjects: Computer simulation; Dynamic heterogeneity of cell populations; Evolutionary simulations; Heterogeneity; Markov model; Nonlinearity; Organisms; Phenotype switching strategies; Stochastic environment; Stochasticity; Strategy; Switching; Therapy; Universality of the statistical characteristics
• ISSN: 0375-9601; 1873-2429
• DOI: 10.1016/j.physleta.2016.01.037

Population heterogeneity plays an important role across many research, as well as the real-world, problems. The population heterogeneity relates to the ability of a population to cope with an environment change (or uncertainty) preventing its extinction. However, this ability is not always desirable as can be exemplified by an intratumor heterogeneity which positively correlates with the development of resistance to therapy. Causation of population heterogeneity is therefore in biology and medicine an intensively studied topic. In this paper the evolution of a specific strategy of population diversification, the phenotype switching, is studied at a conceptual level. The presented simulation model studies evolution of a large population of asexual organisms in a time-varying environment represented by a stochastic Markov process. Each organism disposes with a stochastic or nonlinear deterministic switching strategy realized by discrete-time models with evolvable parameters. We demonstrate that under rapidly varying exogenous conditions organisms operate in the vicinity of the bet-hedging strategy, while the deterministic patterns become relevant as the environmental variations are less frequent. Statistical characterization of the steady state regimes of the populations is done using the Hellinger and Kullback–Leibler functional distances and the Hamming distance. •Relation between phenotype switching and environment is studied.•The Markov chain Monte Carlo based model is developed.•Stochastic and deterministic strategies of phenotype switching are utilized.•Statistical measures of the dynamic heterogeneity reveal universal properties.•The results extend to higher lattice dimensions.