From: Josh Bembenek (jbembe@hotmail.com)
Date: Fri May 09 2003 - 10:30:53 EDT
Jim said:
"The argument concerning the evolution of things of "irrreducible
complexity" seems to embody the notion of a rather linear evolution process.
But the remarkable "processors" of nature are HIGHLY parallel and, taken
with the environments in which they occur, are extremely (unimaginably) rich
and redundant with respect to materials, processes, and instantiations of
things of a given kind and slight variations thereof (e.g., a given
protein). Equally important is the fact that they are also rich with respect
to the huge numbers of potential interactions among them and their products.
The result is essentially EXTREME parallel processing with arguably
exponential outcome possibilities."
-Can you elaborate on this? This seems to be pure imagination. One can
just as easily imagine that all combinatorial interactions of gene products
generates important barriers to system function as one can imagine that all
the combinatorial interactions are favorable and produce further
functionality. There may be a billion fold increase of interactions that
are completely destructive with respect to function compared with those that
increase function in biology. How can a computer performing functional
algorithms answer this? Biology is highly specified and exquisitely
regulated, I don't see this rich potential, at least any evidence of it.
What are you basing this on? Also, in terms of parallel processing, I don't
exactly follow your thoughts here. Are you saying that nature has sampled
all possible combinatorial mutations of all proteins and discovered those
that are functional? I don't fully understand the relationship between
parallel computer processing and rm&ns. For example, the only samples that
ever matter are gametes that are passed on. Perhaps I simply cannot get my
mind around the relevance of these computer models, help me out.
"It would be rather surprising if "processors" and environments such as
these did NOT produce improvements of existing entities and functions, new
entities and functions, and notably new ensemble functions. With respect to
the latter, the second URL reference contained the following observation: "
The crystalline proteins that make up the lens of the eye, for example, are
related to those that serve enzymatic functions unrelated to vision. So, the
theory goes, evolution borrowed an existing protein and used it for a new
function ...that's a lot easier than inventing something entirely new." I
think that nicely summarizes the point of the experimental simulation
results."
-And herein lies the biological problem. Catalyzing reactions and carrying
out biological activities is nothing like completing a computing function.
Whose to tell us the number of possible biological functions for any given
protein? The examples of protein function co-option are few compared to
specific protein families performing specific functions. There's no
justification in saying that sequences can be awarded incremental favorable
selection values, especially prior to them adopting a particular function
(no function equals no selection).
Josh
_________________________________________________________________
The new MSN 8: advanced junk mail protection and 2 months FREE*
http://join.msn.com/?page=features/junkmail
This archive was generated by hypermail 2.1.4 : Fri May 09 2003 - 10:31:13 EDT