Re: design: purposeful or random?

Glenn Morton (grmorton@psyberlink.net)
Wed, 26 Feb 1997 22:59:32 -0600

At 08:50 PM 2/26/97 -0800, Arthur V. Chadwick wrote:
>Glenn originally stated:
>
>>As a point of fact, a randomly chosen sequence has more information in it
>>than an ordered sequence
>
>
>Since our consideration centers on semantics and not on facts, let me attempt
>to define my perspective on information:
>
>1. A random sequence in isolation contains no information whatsoever. It
>has nothing to say about anything,and thus has no information, no inherent
>meaning.
>
>2. A sequence of apparently random numbers may in fact be information-rich.
> For example a portion of the value of pi may appear to be random, but is in
>fact information-rich in its context as a part of the value of pi.
>
>3. Context will always determine whether any sequence of digits or whatever
>is actually information rich. If the random numbers are coming from
>somewhere out there in the solar system, they may in fact be
>information-rich. The same can be said for any measurement of anything.
>But the information is contextual, and without an appropriate cipher, it is
>effectively void of information.
>

This definition of information is not that which is used in information
theory (see below). I would call this "meaning" or a 'semantic form of
information" not "information" in the informational theory sense. But since
this is your post, I will use your definition.

>4. A sequence of bases in a DNA molecule is information rich if when the
>appropriate ciphers are applied, it yields a function related to the
>functioning of some biological organism that contains the DNA.
>
>5. A sequence of random bases in DNA is not information-rich, since it
>does not yield a function related to the functioning of some biological
>organism that contains the DNA
>

You seem to be saying that a sequence which has no ability to to perform a
function has no or little information. I would counter that RNA sequences
are very information rich because a given function seems to be able to be
manufactured in laboratory vats with a frequency of 1 functional molecule
for every 10^14 molecules in randomly made polymers. While this may seem
like a small probability, the large numbers of molecules in one of these
vats almost assures that a functional molecule will be found. (Gerald Joyce
(again) Directed Evolution, Scientific American Dec. 1993) Also Yockey has
shown that 10^93 different cytochromes perform the same function. There is
lots of specificity out there to be found.

>6. Compressability alone (apart from the necessary ciphers) may not be a
>useful criterion for information content.
>

>A useful working definition for information is: data combined with a context
>for meaningful interpretation. While not everyone would agree with this
>definition, I think it is generally in accord with the principles of
>information theory as I understand it.

This would not be the case, according to Yockey. I would point you to
Yockey, Information Theory and Molecular Biology, p. 59-60

He writes,

"There is a great deal of misunderstanding about the relation of the
semantic aspects of a message and the problems of the design and operation
of communication systems, especially by authors who attempt to use
information theory and coding theory or that portion which suits them. I
can't do any better to explain this point than to quote the second paragraph
of Shannon's seminal paper:'The fundamental problem of communication is that
of reproducing at one point either exactly or approximately a message
selected at another point. Frequently the messages have meaning; that is
they refer to or are correlated according to some system with certain
physical or conceptual entities. These semantic aspects of communication
are irrelevant to the engineering problem. The significant aspect is that
the actual message is one selected from a set of possible messages." p. 60

"Although we cannot deal with the semantic --or in molecular biology, the
specificity--aspects of the genetic message, we can establish a measure of
the amount of information in an ensemble of messages." p. 60

Or this,

"Along with many other authors he [Eigen in 1971] makes a play on words by
using 'information' i the sense of knowledge, meaning and specificity. Eigen
& Schuster (1977) use 'information' in the correct sense according to
eqation (2.24) but in the same section of their paper the use 'information'
in a semantic sense." p. 272

Your definition makes 'information' be the semantic version.

glenn

Foundation, Fall and Flood
http://www.isource.net/~grmorton/dmd.htm