QUESTIONS: alpha-helix "signals" in proteins

[Joshua Cherry]

In article <Pine.3.89.9407121006.A5334-0100000 at fox.cce.usp.br> szeinfel at FOX.CCE.USP.BR (Rafael N Szeinfeld) writes:
   
	   I'd like to say that there are two diferent problems behind the
   protein folding problem.  
	   First how proteins solve the Levinthal paradox that is, how a
   protein folds so fast with so many configurations to be tested in order to
   find out the energy minimun (this minimun being or not path dependent). 
   This seems to be sequence independent since all proteins folds fast (in 
   comparison to the time it would take to search the whole configurational 
   space).

I wouldn't say that fast folding MUST be sequence independent just 
because "all" proteins fold fast.  After all, proteins are the
products of selection, and that selection requires reasonably
fast folding.  You wouldn't want to say that because all proteins
fold, the property of folding is sequence independent.
	The interesting question is whether in fact most sequences
with a thermodynamically stable folded state will fold reasonably
quickly.  I like to think about the problem in terms of all
possible sequences of a given size range.  Of this set, some
subset (I would guess small) will have a thermodynamically 
stable structure which we would call a folded state.  One question
is: what fraction of this subset would actually assume its
folded state in a "reasonable" amount of time?  If the answer to this
is that 90% would do so, I would say that the kinetics of folding
are maybe not so interesting.  If the answer is one in a million,
then fast folding is a very special property of certain sequences,
and is clearly worth studying.  I am truly agnostic on this 
issue.

Josh Cherry
cherry at watneys.med.utah.edu