Newsgroups: comp.ai.alife
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!howland.reston.ans.net!pipex!uunet!timbuk.cray.com!walter.cray.com!mwd
From: mwd@cray.com (Mark Dalton)
Subject: Re: "recombinant mutations" + selection --> rapid evolution
Message-ID: <1995Jan14.163651.6360@walter.cray.com>
Lines: 104
Nntp-Posting-Host: pajarito.cray.com
X-Newsreader: TIN [version 1.2 021193BETA PL3-CRIb]
References: <3f3f0p$6kn@ixnews2.ix.netcom.com> <3f4ap8$fnr@scapa.cs.ualberta.ca>
Date: 14 Jan 95 16:36:51 CST

I am sorry, I did not read the articles yet, but here are some general
thoughts on GA in comparison to genetics vs. evolutionary theory (versus
biological evolution).

Terry Van Belle (vanbelle@cs.ualberta.ca) wrote:
: Looks like an interesting article, but it seems to have a few
: misconceptions about evolution.  First of all, both point mutation and
: recombination (at least as Holland defines it) are random processes.

When done in organisms the more random, the more offspring and natural
selection plays a higher role.  As the organisms get more complex the
processes are less random.

	Also a side note.  General 'less randomness', with UV light that
will affect the T's of DNA, causing to T's to bind together, causing
the misread.  Non-coding regions have the highest concentration of A's and
T's.  So, for UV's affect on DNA the odds are for all organisms that it
is less likely to cause a mutation that will affect the organism.

It is true that Genetics (especially as the organisms get more complex),
mutations and recombinations are more rare or unlikely to occur, especially
in the germ-line cells.
   For Multi-cellular organisms because of:
	1. Mutagens accessbility to germ-line or dividing cells.
	    - Most UV light only gets to the upper layers of skin a mutation
	      usually is not important because most of those cells do not
	      divide and will die soon.
	2. Many of the cells don't divide or have a limited number of
	   divisions.  (one thing cancer does is causes increased cell
	   division).
	3. Repair/error checking mechanisms of 'higher' cells.
	4. Packing of genetic material and interior folds.  This will reduce
	   certain segments of DNA accessiblity and likelihood of mutation.
	5. Double stranded/complimentary DNA.  

     - Recombination still can occur or read errors for mutations, but the
	affect of many of the causes of mis-reads are less _likely_.

   For single cell organisms recombination and mutation are more likely:
	1. The cell is the 'germ-line' cell - offspring are derived from
	   its Genetic material (mutations and all).
	2. Higher exposure to mutagens (UV, chemicals, etc.).
	3. less repair/error checking mechanisms.
	4. The genetic material, even more in prokaryotic (no nucleus)
	   cells, are susceptable to mutations, and have less folding
	   and protection.  Even exposed circular DNA or plasmids which
	   are exchanged between cells.  (These plasmids may contain
	   enzyme for a specific function).  (see conjugation, replicons,
	   transposons).
	5. The number of offspring is much higher, thus more changes can
	   occur over the same time.

: I'm not sure what's involved in preselecting segments, but the classic
: GA, like evolution, only uses fitness information for the whole
: genome.  This, however, is beside the point.  While mutation and
: recombination may be random processes, they are only half of the
: evolution equation.  The other half, natural selection, is decidedly
: non-random.  Thus Darwinism does not, and never has postulated that
: the myriad wonders of nature are the result of a random process.
: It is true that genetic algorithms tend to converge to solutions far
: more rapidly if they use recombination, but this fact merely
: illustrates the superiority of sexual reproduction over asexual,
: hardly an earth-shattering conclusion.

Cool stuff that make sense for multicellular recombinations:
	1. Antibody production - there are the heavy and light chain
		protein recombinations.
	2. In Myasthenia Gravis - we noticed something that may support
	   this in that, there seemed to be functional units that kept
	   showing up in aprox. units of 9 amino acids (27 Nucleic acids).
	3. Jumping Genes.

Other stuff:
	1. Higher organisms have the advantage of a chromosome from each
	   parent, allowing dominant and recessive traits, backup copies
	   of a gene, and/or slight modifications.
	2. Genetic Algorithms and programming tend to focus on 'genetics'
	   from a evolutionary perspective, rather than a biological
	   perspective  (ie. mutation,recombinations,etc. are all miss
	   many of the control,modification, mechanisms).  Also you
	   don't go straight from DNA --> function in biology.
	    - There is viability of the RNA, specific sequences must
	      be maintained others not, structural similarities for proteins
	      and RNA can be VERY important.  Not to mention that the
	      DNA is translated to RNA which is translated to protein
	      which then is used for a function.  (DNA/RNA --> protein
	      allows for some modification most 'safe' on the last
	      Nucleic Acid of the codon).
	
In general:
	The functional perspective has been greatly under studied, due
to the assumption that it must be a relation if similar (as a first thought),
instead of objectivity.

Mark
----------
Mark Dalton       CH3-S-CH2 H                      H      O       H
Cray Research,Inc.      |   |                      |       \      |
Los Alamos,NM 87544     CH2-C-COO    //\ ---C--CH2-C-COO    C-CH2-C-COO
mwd@cray.com                |       |  ||   ||     |       //     |
                            NH3      \\/ \ / CH    NH3    O       NH3
                                          NH
URL = http://lenti.med.umn.edu/~mwd/mwd.html

