Re: Functional proteins from a random library

From: Paul Nelson (
Date: Fri Apr 06 2001 - 05:40:03 EDT

  • Next message: george murphy: "Re: ID.Weinberg"

    Preston Garrison wrote:

    > For those who might be interested in the question of
    > how rare are functional protein sequences among the
    > huge number of possible sequences, there is a
    > breakthrough paper in this week's Nature. The authors
    > have solved a number of technical problems to create
    > a library of ~10^13 mRNAs encoding proteins 80
    > amino acids long. This is a much larger library than is
    > obtainable by other approaches in this field. They
    > have arranged it so that after in vitro translation
    > into protein, the mRNA and protein remain connected
    > to each other, enabling the association of a protein
    > function with an amplifyable RNA sequence. Using this
    > system they found 4 sequence classes of ATP binding
    > proteins in their library, none of which seems to be
    > related to known biological sequences. The highest
    > affinity binding was 100 nM, which is quite respectable.
    > This is only the beginning of what should be very
    > interesting results from this system. If it turns out
    > to be possible to get lots of functional proteins out
    > of libraries this size, it will not be good for
    > Mr. Dembski's argument.

    Bill Dembski deals with experiments like this (as I
    recall, he looks at ribozyme engineering) in his
    forthcoming book, _No Free Lunch: Why Specified
    Complexity Cannot Be Purchased without Intelligence_.

    Pools of ~10^13 mRNAs could not exist on the early
    Earth, or anywhere, really, without the help either
    of organisms or clever biochemists. Indeed it is
    possible to "back out" of experiments like these
    one step at a time, removing the intelligently-
    synthesized reagents (e.g., buffers) and artificial
    conditions. [In some ribozyme engineering
    experiments, for instance, the RNAs are tethered
    to keep them from precipitating.] As intelligent
    control or intervention is removed, the results
    cease to be biologically relevant.

    Bill's point exactly, as he explains in _No Free
    Lunch_. A library of ~10^13 mRNAs is as designed
    an object as an integrated circuit. To put it another
    way, you can be sure that Jack Szostak and colleagues
    don't start their experiments under prebiotically
    plausible conditions. No mRNA by that route. ;-)

    Paul Nelson
    Senior Fellow
    The Discovery Institute

    This archive was generated by hypermail 2b29 : Fri Apr 06 2001 - 05:38:43 EDT