Saturday, July 12, 2008

The Fluctuation Test

John Dennehy of The Evilutionary Biologist continues his almost perfect1 record of picking important papers for his Citation Classic.

This week's paper is the classic 1943 paper by Luria and Delbrück on the fluctuation test [The Fluctuation test]. This was the paper that proved that mutations arise randomly with respect to their phenotype. As John says, it is one of the most important experiments in biology.

1. For the exception see It Happens to all of Us Eventually.


  1. I learned about this last semester in my microbial genomes class. An important experiment and an easy experiment to understand I think.

  2. Here is the link to John's post. I hope some of the posts from his series make it to the new carnival - the Giant's Shoulders.

  3. An ultimate test of lamarckism?
    Pffft. That's...dumb.

    The only thing they show is that negative selection cannot produce direct adaptive mutations. There is not even the environmental induction of a new phenotype involved here

    See Agrawal et al 1999 for a true testing of inheritance of acquired traits, with positive results...or Cropley et al. 2006 PNAS paper for an experimental case in the mouse (involving mehtylation)

  4. Genetic assimilation of a trait is also possible by artificial selection, see Waddington's experiments with Crossveinless.
    Also, ether is enough (without selection) to produce and assimilate the Ubx phenotype in drosophila. My guess is parallel methylation of both somatic and germ lines, possibly reinforced by succesive generations of exposure.

  5. Umm, genetic assimilation isn't lamarckism, its just selection on phenotypic plasticity. I'd say maternal effects are a better example of inheritance of a parent's environment.

  6. the tension between chance and necessity, between randomness and repeatability

    The quote from Lenski ties directly back to your post on Monod. Seems like chance, as used by Lenski, is actually contingency as opposed to, say, reproducible stochastic distributions.

    Confusing, but hopefully it makes sense when reading Monod.

  7. Actually, Martin, not quite. Waddinton only eliminates those who fail to modify phenotype, but is ot selecting for lower threshold or anything like that. Ypu may call tehes "genes of phenotypici plasticity" (whatever THAT is supposed to mean). I think these ar juts genes that have effects similar to the environmental stimulus and can become accumulated

    Waddington experiment, unlike the fluctuation test, qualifies as the inheritance of an acquired trait because in his initial population flies need the environmental stimulus to develop the phenotype, and he ends up with a population where most flies develop the phenotype without the stimulus.

  8. The fluctuation test was a bit of pure elegance. I love papers like that. The Lederberg's replica plating experiment nailed it too, IMHO.

  9. More like a bit of pure ignorance, if you think it has anything to do with inheritance of acquired traits. You would have to, first, induce a modified phenotype environmentally, and then see if you still obtain that phenotype in the progeny thereafter, despite absence of exposure to the environmental stimulus. That is the standard "lamarckian" experiment.

  10. The fluctuation test was a bit of pure elegance. I love papers like that.

    Yes indeed! Even if haven't read it, I look at the figures and go "hot damn!"

    And I'm not even in the biz.

  11. Randomness i the mergence of an adpative mutaion...that's cool.
    Utimate Refutation of lamarckism...that's stupid.