Gene networks accelerate evolution by fitness landscape learning

John Reinitz, Sergey Vakulenko, Dima Grigoriev, and Andreas Weber
Sept. 2016
 

Abstract

We consider evolution of a large population, where fitness of each organism is defined by many phenotypical traits. These traits result from expression of many genes. We propose a new model of gene regulation, where gene expression is controlled by a gene network with a threshold mechanism and there is a feedback between that threshold and gene expression. We show that this regulation is very powerful: depending on parameters we can obtain any functional connection between thresholds and genes. Under general assumptions on fitness we prove that such model organisms are capable, to some extent, to recognize the fitness landscape. That fitness landscape learning sharply reduces the number of mutations necessary for adaptation and thus accelerates of evolution. Moreover, this learning increases phenotype robustness with respect to mutations. However, this acceleration leads to an additional risk since learning procedure can produce errors. Finally evolution acceleration reminds races on a rugged highway: when you speed up, you have more chances to crash. These results explain recent experimental data on anticipation of environment changes by some organisms.

Bibtex

@MISC{ReinitzVakulenkoGrigorievWeber2016a,
    author = {Reinitz, John and Vakulenko, Sergey and Grigoriev, Dima and Weber, Andreas},
     title = {Gene networks accelerate evolution by fitness landscape learning},
      year = {2016},
     month = sep,
  abstract = {We consider evolution of a large population, where fitness of each organism is defined by many
              phenotypical traits. These traits result from expression of many genes. We propose a new model of
              gene regulation, where gene expression is controlled by a gene network with a threshold mechanism
              and there is a feedback between that threshold and gene expression. We show that this regulation is
              very powerful: depending on parameters we can obtain any functional connection between thresholds
              and genes. Under general assumptions on fitness we prove that such model organisms are capable, to
              some extent, to recognize the fitness landscape. That fitness landscape learning sharply reduces the
              number of mutations necessary for adaptation and thus accelerates of evolution. Moreover, this
              learning increases phenotype robustness with respect to mutations. However, this acceleration leads
              to an additional risk since learning procedure can produce errors. Finally evolution acceleration
              reminds races on a rugged highway: when you speed up, you have more chances to crash. These results
              explain recent experimental data on anticipation of environment changes by some organisms.},
       url = {https://arxiv.org/abs/1609.08784}
}