protein / predicting 3D-structure perturbation

Scott Le Grand legrand at teslaNOSPAM.mbi.uclaFNORD.edu
Fri Sep 12 22:53:21 EST 1997


Marketa J Zvelebil wrote:
> It would depend what kind and how large are the perturbations you
> want to measure. Easiest is to model the mutation as you have and
> simply minimize the structure and look at local changes. If you want
> to go further you can try local dynamics. If you really want to get
> deeply into it, with probably very little extra information, you
> could try things like how has it effected the charge distribution
> around that residue etc. That is all available within insightII
> (Biosym).

Unfortunately, the theoretical models behind insightII and all
other force fields out there bear no known resemblance to reality.
An energy minimization will do wonders with removing steric
conflicts as a result of the mutation, but it will do nothing
to assist you in the prediction of genuine structual changes
caused by the mutation.

The recent CASP and CASP II structure prediction competitions provided
the first real opportunities to test these models objectively on
blind data.  No one came out a winner except in their own mind and
that's the hard pill we all ought to swallow, but many won't.

Scott Le Grand, Ph. D.




More information about the Bio-soft mailing list