Re: Functional proteins from a random library

From: Paul Nelson (
Date: Mon Apr 09 2001 - 11:15:18 EDT

  • Next message: Loren Haarsma: "Re: Functional proteins from a random library"

    Preston Garrison wrote:

    > Neither I nor the authors claimed that this work
    > had anything to do with the origin of life.
    >We use synthesized reagants and artificial conditions
    >to do all of biochemistry and genetics. Does that mean
    >that biochemistry and genetics are biologically irrelevant?
    >If intelligent control invalidates an experiment in this
    >area, then _all_ experiments are invalidated.

    Nonsense. Origin-of-life investigators routinely
    debate the merits of each other's experiments in
    terms of their prebiotic plausibility. I do not
    think you would credit an experiment claiming to
    show the natural or prebiotic (i.e., undirected)
    synthesis of ribozymes, for instance, if the
    investigators began with RNA from Sigma.

    Experiments can be more, or less, illuminating
    of any particular problem, depending on how
    closely we take their conditions to mirror nature.
    Of course experiments are "intelligently-designed,"
    in the sense that we arrange what Francis Bacon
    called "a trial," to understand how nature herself
    operates. But if a *result* depends critically on
    the intervention of an agent -- i.e., if we see that
    nature on her own would not achieve the same
    -- then it is willfully naive to pretend that we
    didn't intervene.

    >>A whole lot of specified information, provided by
    >>Szostak and his co-workers, is needed even to
    >>reach the point where one can go fishing for functions
    >>in large pools of mRNAs (and their products).

    >This is utterly irrelevant to the question of where the
    >sequence information came from. This is again an argument,
    >in effect, that all experiments are invalid. No sequence
    >information went into the library (or very little - there
    >is probably some weak bias from aspects of the library
    >construction). Sequences came out that bind ATP. The
    >sequence information (the only kind of information that
    >is relevant to interpreting the experiment) was not put
    >in by the investigators.
    >The experiments are designed to begin answering the
    >specific question, "How rare are foldable/functional
    >protein sequences in sequence space?" This is an
    >interesting question from the standpoint of protein
    >biophysics alone, but it is obviously relevant to the
    >question of whether and how completely new proteins
    >can arise in non-functional or frame shifted sequences
    >of existing organisms.
    >The only reason I alluded to Bill Dembski in particular
    >in my message is that I gather that, in contrast to Behe,
    >he believes that even individual proteins cannot arise
    >by chance.

    In a prebiotic context: yes. Nothing in the Nature paper
    challenges that.

    From my reading of Szostak's work, his central goal is
    to solve the problem of the natural (undirected) origin
    of life. I'll ask him in a couple of weeks about this
    recent paper, in relation to his long-term research
    goals. I withhold judgment on the *biological*
    relevance of these new experiments, but I reaffirm
    my opinion that their prebiotic relevance is nil.

    >The number I remember is 10^60 for the odds that he
    >gives, which is too rare to find by chance. The results
    >of these experiments begin to suggest that he is wrong
    >about that one point, but obviously much remains to
    >be done to see how rare different kinds of functions are.

    The number Bill estimates refers to a prebiotic context.

    >As it happens, I am sympathetic to the ID position,
    >especially at the point of the origin of life, since I
    >don't see any way that that nut can be cracked. It may
    >be that, to Howard's and George's theological chagrin,
    >the Lord did miracles along the way to get life going,
    >and perhaps at points after that. But no one in science
    >is going to listen to any argument from a messenger who
    >comes also making the remarkably bad argument that an
    >experiment proves nothing simply because it was designed
    >(i.e., that it WAS an experiment.)

    As I said, experiments can be more, or less, illuminating
    of any particular problem depending on their closeness
    of fit to nature. My point was not that Szostak et al.
    did an experiment. It was, rather, that its relevance to
    the origin of life is doubtful. On that we appear to

    Paul Nelson

    This archive was generated by hypermail 2b29 : Mon Apr 09 2001 - 11:13:57 EDT