Computer-generated vaccines

Tracy P. Hamilton chem013 at uabdpo.dpo.uab.edu
Mon Nov 22 16:45:16 EST 1999


Pierre wrote in message <818dle$2jee$1 at news4.isdnet.net>...
>What is the role of contingency/probability in protein folding ? If one
>chain has many possible foldings, each affected with a given probability
>amongst the population of molecules (like in evolution/population
genetics),
>is it possible to calculate all the variations ?


[snip my previous post]

One would not be able to go thorough all possible protein conformations.
I *believe* (I am no protein folding expert, but a theoretician in a more
fundamental sense) that current approaches are more empirical.
Predict secondary structure (alpha helix, beta sheet, random coil) from
part of the amino acid sequence, tertiary structure (overall shape) is from
a different database, but similarly empirical.  Both are correct about half
the time now, last I heard.

A group was able to predict the sequence of a protein that would be
the same shape as one of the zinc finger proteins, except without zinc!
Makes one wonder why zinc fingers have zinc...

Theory does currently play a large role in constraining structures,
particularly
NMR ones (solution structures based on H-H distances).  I don't have a
good feel for what would happen if you applied one of the good protein
force fields to the superoxide dismutase enzyme (or the mutant that
has been identified with inherited ALS which has a more exposed
active site).

If I were to model the active site, I would probably put some geometric
constraints on the active site to keep it in the neighborhood for at least
some calculations.

Tracy P. Hamilton






More information about the Immuno mailing list