fec1375d.0212051446.528bb0f4 at posting.google.com>
I'd like to throw some thoughts to stimulate a discussion.
Protein folding has long attracted the interests of theoretical
and
computational physicists, as well as other types of
theoretically-minded people.
There has been attempts to say something about folding based on
concepts inspired from spin glass theory, ``energy landscapes'',
and
a plethora of statistical mechanical and other models. What did
this accomplish?
Not much, it seems to me. The spin glass analogy has proven
pretty
much useless to understand what proteins really do. The ``new
view'',
namely that folding is driven by a ``funnel''energy
landscape, is quite obvious from a physical point of view and
doesn't add anything really useful (to make predictions, that
is).
Indeed, the basic physics (i.e. the physics which is common to
all
proteins) is quite simple.
Despite this, the hype around protein folding physics continues
unabated. I would predict that at some points the funding
agencies and
the public will realize that there isn't much beef here, and at
that
point the whole field of energy landscape-ology etc. will be
quickly
forgotten.
The real, important scientific issues in protein folding have to
do to
the SPECIFIC behavior of a given protein and its interactions
with
other molecules. The devil really is in the details here. But
this is more
an experimental problem than anything else. One can hope that at
some points people will be able to do detailed, realistic
simulations.
Whether these can provide insight that is not obtainable from
experiments remains to be seen, in my view.
