Advertisement
Research Article

Natural Selection Fails to Optimize Mutation Rates for Long-Term Adaptation on Rugged Fitness Landscapes

  • Jeff Clune mail,

    jclune@msu.edu

    Affiliations: Department of Computer Science and Engineering, Michigan State University, East Lansing, Michigan, United States of America, Instituto de Biología Molecular y Celular de Plantas, Consejo Superior de Investigaciones Cientificas, Universidad Politecnica de València, València, Spain

    X
  • Dusan Misevic,

    Affiliation: Institute of Integrative Biology, ETH Zurich, Zurich, Switzerland

    X
  • Charles Ofria,

    Affiliation: Department of Computer Science and Engineering, Michigan State University, East Lansing, Michigan, United States of America

    X
  • Richard E. Lenski,

    Affiliation: Department of Microbiology and Molecular Genetics, Michigan State University, East Lansing, Michigan, United States of America

    X
  • Santiago F. Elena,

    Affiliations: Instituto de Biología Molecular y Celular de Plantas, Consejo Superior de Investigaciones Cientificas, Universidad Politecnica de València, València, Spain, The Santa Fe Institute, Santa Fe, New Mexico, United States of America

    X
  • Rafael Sanjuán

    Affiliations: Instituto de Biología Molecular y Celular de Plantas, Consejo Superior de Investigaciones Cientificas, Universidad Politecnica de València, València, Spain, Institut Cavanilles de Biodiversitat i Biologia Evolutiva, Universitat de València, València, Spain

    X
  • Published: September 26, 2008
  • DOI: 10.1371/journal.pcbi.1000187

Reader Comments (4)

Post a new comment on this article

Fidelity of mutator locus constant or variable?

Posted by bostman on 30 Sep 2008 at 20:53 GMT

From your descriptions I am unsure whether the fidelity of \mu was held constant or if it was variable.

From the results section page 3:
"Finally, we let the mutation rate apply reflexively to itself, such that high-fidelity genotypes rarely changed their mutation rates whereas low-fidelity genotypes did so frequently."

But in the methods section page 7 you say:
"In treatments where the mutation rate could change, \mu
had a constant and high probability \Pi of changing by a small
amount during any replication cycle."

Which is it?


Both!

bostman replied to bostman on 01 Oct 2008 at 07:09 GMT

Ok ok, I missed it in the first reading. Sorry. You did both, of course. And I see no matter what, the mutation rates evolved to suboptimal rates.

I've done a similar experiment in the NK-landscape, where the K parameter controls the ruggedness. For a rugged landscape (high K), I found that I could only evolve non-zero mutation rates if I reset the landscape at random every ten generations.