Get access

THE CAUSES OF EPISTASIS IN GENETIC NETWORKS

Authors

  • Javier Macía,

    1. Complex Systems Laboratory, ICREA-Universitat Pompeu Fabra, 08003 Barcelona, Spain
    2. Instituto de Biología Evolutiva, CSIC-Universitat Pompeu Fabra, 08003 Barcelona, Spain
    Search for more papers by this author
  • Ricard V. Solé,

    1. Complex Systems Laboratory, ICREA-Universitat Pompeu Fabra, 08003 Barcelona, Spain
    2. Instituto de Biología Evolutiva, CSIC-Universitat Pompeu Fabra, 08003 Barcelona, Spain
    3. The Santa Fe Institute, Santa Fe, New Mexico 87501
    Search for more papers by this author
  • Santiago F. Elena

    1. The Santa Fe Institute, Santa Fe, New Mexico 87501
    2. Instituto de Biología Molecular y Celular de Plantas, CSIC-Universidad Politécnica de Valencia, 46022 València, Spain
    3. E-mail: sfelena@ibmcp.upv.es
    Search for more papers by this author

Abstract

Epistasis refers to the nonadditive interactions between genes in determining phenotypes. Considerable efforts have shown that, even for a given organism, epistasis may vary both in intensity and sign. Recent comparative studies supported that the overall sign of epistasis switches from positive to negative as the complexity of an organism increases, and it has been hypothesized that this change shall be a consequence of the underlying gene network properties. Why should this be the case? What characteristics of genetic networks determine the sign of epistasis? Here we show, by evolving genetic networks that differ in their complexity and robustness against perturbations but that perform the same tasks, that robustness increased with complexity and that epistasis was positive for small nonrobust networks but negative for large robust ones. Our results indicate that robustness and negative epistasis emerge as a consequence of the existence of redundant elements in regulatory structures of genetic networks and that the correlation between complexity and epistasis is a byproduct of such redundancy, allowing for the decoupling of epistasis from the underlying network complexity.

Get access to the full text of this article

Ancillary