Re: Is there any reason for the evolution to be one way
- From: "Perplexed in Peoria" <jimmenegay@xxxxxxxxxxxxx>
- Date: Tue, 27 Mar 2007 00:01:16 -0500 (EST)
"Shaktyai" <Fabrice.Allais@xxxxxxxxx> wrote in message news:eu7l0b$1rlj$1@xxxxxxxxxxxxxxxxxxxxxx
Random walk with p=q=0.5 have a zero mean deviation: ie no evolution.
That goes for a model where a mutation of a gene is modeled in a 1D
space. One step on the right or one step for each mutation from state
A to B, from B to C, etc...
Such a model can not reproduce the experimental facts: evolution =
there is a non zero mean deviation = there is a drift in the random
walk.
Yes, such a model can reproduce the experimental facts. Evolution happens
without a mutation bias, due to selection. Evolution can also happen with
neither a mutation nor a selection bias, in which case it is called drift.
Less well known and less important than either of these is evolution due
to a mutation bias. If you want to know about this, I have suggested that
you Google on
mutation pressure G+C
But you need to know that this has nothing to do with why most evolution
takes place.
I have in mind the 1D model (and I may be wrong there !), so the
only solution is to assume that in general: p(A->B) is different from
p(B->A).
To justify this I was thinking about the thermodynamics of a mutation.
In my point of view, the two probability can be equal if E(A)=E(B).
If the free energy of the initial and final state are different then
the system will always spontaneously evolve toward the lower energy
state and the two probability should be different. Of course there is
always the possibity to bridge the energy gap with either a chemical
catalyst, a photoabsortion, or an electronic colisionnal excitation
(electron beam).
I might be all wrong but is there a systematic study of the energy
required for a mutation to spontaneously happen ?
There is a lot of talk about Hamiltonian of Darwin's systems.
Not among biologists, there isn't.
To write
down a satisfactory theory, one must have a density probability whose
integral is always one. In kinetics, this very fundamental point is
induced by the hamiltonian structure that induces Liouville's theorem.
There is nothing corresponding to Lioville's theorem in biology. Biology
does not involve any useful conservation laws. The basic units in biology
*reproduce*. That doesn't happen in chemistry or physics.
In information theory, the probility space is constant and well
defined. But in biology it is far from obvious.
You need to learn something about biology before you try to apply your
physicist's intuitions to the subject. You will no doubt learn that
biology is far more difficult and almost as sophisticated mathematically.
Since you will probably want a mathematically oriented treatment of
evolution, I would suggest that you start your education with Sean Rice's
book "Evolutionary Theory: Mathematical and Conceptual Foundations".
Also, read the chapters on mutation in the Alberts textbook on Molecular
Biology.
.
- References:
- Is there any reason for the evolution to be one way
- From: Shaktyai
- Re: Is there any reason for the evolution to be one way
- From: Shaktyai
- Is there any reason for the evolution to be one way
- Prev by Date: Re: Is there any reason for the evolution to be one way
- Next by Date: Re: Is there any reason for the evolution to be one way
- Previous by thread: Re: Is there any reason for the evolution to be one way
- Next by thread: Re: Is there any reason for the evolution to be one way
- Index(es):
Relevant Pages
|
