[gmx-users] parameters

[Nagy, Peter I.]

Dear Gromacs Users,

   I am performing DFT calculations for obtaining parameters for the acylated cystein:

   Cv-Cu-C(=O)-S-Cb-Ca

I need torsion potentials for Cv-Cu-C-S, O=C-S-Cb, and C-S-Cb-Ca. (I want to see the deviation of the latter from that used in methionine with sp3 carbon.)
Thus I have performed torsion potential calculations about bonds Cu-C, C-S, and S-Cb. 
  
   The first question: when I perform rotations about the Cu-C bond, it describes an overall conformational change. How can I assign it to the Cv-Cu-C-S torsion, while it also may characterize the Cv-Cu-C=O torsional
potential. In the calculations I set only the torsional angle to some specific
values, and optimized the rest of the molecule.

   Second question: in obtaining the C-S bond stretching force constant, (and the C-S-Cb bond angle force constant in another set of calculations) I am doing calculations with different C-S distances. I have the optimized structure with full geometry optimization. Then what is the procedure consistent with the Gromacs parameterization: should I perform single point
calculations with modified C-S distances, or should I set different C-S distances and optimize the rest of the molecule. The latter procedure should
result a stretching potential more relaxed, where all internal coordinates
follow the change of the C-S distance, thus it would be an "adiabatic" parameter. Which one is the consistent one? In the manual Figs. 6,7 shows a
parabolic energy function. I would use such a fit, but the question is:
are the energies obtained from a single-point set or from relative energies with relaxed structures.

Thanks, Peter Nagy

pnagy at utnet.utoledo.edu

Dept. Med. Biol. CHem.
The University of Toledo
Toledo, OH 43606
C-S distances

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20031223/e6fd6aad/attachment.html>