From: Peter Passaro (ocsrazor@worldnet.att.net)
Date: Mon Sep 21 1998 - 17:59:15 MDT
John,
I have been following your discussion, and as a practicing molecular
biologist, I couldn't resist helping you clarify some key points:
The gene you have been discussing is a so called master regulator gene.
Genes in this category function during the development of organism to call
for cascades of cell differentiation and structural reorganization. In
this way they are interesting for the points you have mentioned, but, to
date, there is nothing that would indicate that there is a higher level
code with possibly the exception of protein subunits (more on that below).
Genes like this are simply switches which turn a downstream event on or
off. You could use this gene to control many developmental pathways if you
were able to get it to "talk" to the downstream signal transduction
machinery that is already in place in an organism.
> >A code allows you to decode something.
>
> Yes and even more important in this case, a high level code would give
you an
> easy way to encode a complex physical trait into DNA.
The codes you are looking for have already been found. Other than the
codon triplet of nucleotides in DNA, there are the subunits of proteins
found in signal trandsduction cascades. By cataloging the functions of all
known protein subunits we could possibly save the time of designing a
protein from scratch. By reassembling subunits you could produce a protein
with new properties, or change the direction in which signals move along a
pathway in a cell.
> >Knowing the eye-signal gene allows you to spot other eye-signals
> >but gives you no idea what any other gene does.
>
> I said there was good evidence a high level language must exist, I don't
know
> how to reach any other conclusion when we find a simple command can cause
> thousands of mouse genes to cooperate and produce a very complex
structure
> like a mouse eye and exactly the same command can make thousands of fly
genes
> produce a vastly different fly eye. I also made very clear that we don't
> understand that high level language yet, we only know two or three of the
> words and none of the grammar. My point was that interesting things could
> result when we know more and we may learn more in a hurry.
The eye signal along with thousands of other developmental regulation
signals are highly conserved across all animals from worms on up. The
reason for this seems to be that the cambrian radiation of animal species
from which all veterbrates and invertebrates arose seems to have happened
in an eyeblink of geologic time, so, fortunately for molecular biologists,
there wasn't enough time once the core genetic mechanisms were set up to
deviate through mutation. If a creature deviated too far from the
entrenched mechanisms it died.
> >With the genetic code, by contrast, I can give you the amino acid
> >sequence of any gene
>
> Genes do not contain amino acids, proteins do. Genes are made of DNA
> (RNA in a few viruses like AIDS) and information is conveyed in a
> very long sequence of only 4 bases in that DNA. The genetic code is just
> the mapping of the 64 possible base triplets (4 X 4 X 4) to the
> 20 amino acids life uses in proteins.
You misunderstood his point. If you have the genetic code you also have the
amino acid code of every protein in an organism.
> >not different in any functional sense.
>
> Exactly my point. Call it a signal, call it a code, call it a language,
> whatever you call it one thing is certain, it itself is not an eye, for
it to
> produce an eye a translation is needed. The amazing thing is that all
life
> seems to use the same translation book, and that's why I want a copy of
that
> book.
As I mentioned above, protein subunits are the only thing that looks like a
higher code. There is a feeling in molecular biology that once we have a
great deal of knowledge of the different types of subunits, we could
recombine them or design new ones (nanotech starting with proteins). But
what this does not give you is the design for better large scale
structures. Downstream of a master regulator gene there is a dizzying
array of signaling machinery which appears to have been assembled largely
through random selection over millions of years. If we were to design
large scale biological structures we would have to deal with a number a
variables that is staggering, not impossible, but much guidance will have
to come from the fields of complexity and chaos before something like this
is even thinkable.
> >A human who had all the highly conserved genes from a chimp would be
> >a human to all tests, except to a sequencing machine.
>
> As written that's pure unadulterated nonsense. My hope is that's not
> really the idea you want to express and you just wrote a bad sentence.
He is absolutely right! The reason that we have better than a ninety eight
percent genetic similarity with chimps is that all of our low level code is
nearly identical. Where we differ is some key internal and external
structural differences. Those differences probably depend on something
like a gene duplication event or a rearrangement which produced new
functions which were deleterious to the organism as a whole. BUT, these
changes are the result of changing minute variables in an incredibly
complex system which has been assembled at random on top of other
incredibly complex systems which were also assembled at random.
Don't get me wrong I love biology, it is one of my true passions. We have
learned a tremendous amount about complex systems and molecular design from
observing nature. But mother nature is not efficient by any stretch of the
imagination. I'll put my faith in this incredibly complex system sitting
on top my shoulders to bootstrap us to systems which can understand the
level of complexity on which nature works. Right now, I just don't think
we have the processing power. In other words, my money is on the
neurosciences, AI, and nanotechnology to produce the kinds of redesign we
are discussing. Unfortunately molecular biology is a great teacher, by I
don't think it will be the instrument of change
This archive was generated by hypermail 2.1.5 : Fri Nov 01 2002 - 14:49:36 MST