Hello Violet,
I do not know if you tried to use a template or not for your
pentapeptide modeling. In case not, the set deviation only sets the
starting model atoms at a different random position, then typical angle/
dihedral angle preferences, as derived from statistics, will dominate
your model building: likely to produce some kind of extended
conformation, because that is the one that avoids the best any spatial
conflict. If there is a template, then additionally, there are distance
and angle dih angel preferences derived from the template. They will
dominate your final conformation, the "set deviation" will introduce
only small perturbations into your final model.
for a pentapeptide 1.7 A RMS is pretty large, I think it is essentially
a random sampling within some reasonable perimeter.
Andras
Violet Chang wrote:
>
> Hi Modellers,
> I understand that it's not a good idea using Modeller to model the very
> small peptdie(Here I tried to model the pentapeptide). And I've read the
> reply posted by Fiser about forcing modeling to deal only with sidechains
> and retain the same backbone conformation. However, I'd like to know if
> it's possible to achive the same goal by setting the "SET DEVIATION =
> 0.0'.
>
> I have built different models by different .top files. The first one is
> with 'SET DEVIATION= 0.0'. And the second one, in addition to 'SET
> DEVIATION= 5.0', is included a subroutine called 'select_atoms'. Then the
> rms among models and templates are around 0.435-1.774. Can you please help
> me explain the results? I don't know how much meanings my models have.
>
> Thanks very much.
> Best regards,
> Violet
>
> Violet(Chiung-wen) Chang, Ph.D. Student
> Department of Physiology and Biophysics and Institute for
> Computational Biomedicine, Mount Sinai School of Medicine
> One Gustave L. Levy Place, Box 1022, New York, NY 10029
> phone: (212) 241 1611
--
,
Andras Fiser, PhD # phone: (212) 327 7216
The Rockefeller University # fax: (212) 327 7540
Box 270, 1230 York Avenue # e-mail:
New York, NY 10021-6399, USA # http://salilab.org/~andras