On Thu, Feb 21, 2013 at 2:13 AM, DEBOSTUTI GHOSHDASTIDAR <
debostutighosh.gmail.com> wrote:
> Dear Amber Users...
>
> I am using OPLS-AA force field parameters for running simulations for ionic
> liquid using Amber. I have some doubts regarding the preparation of the
> dihedral part of the frc file using these OPLS parameters. I would use an
> example to explain the doubt.
>
> >From the paper JCTC (vol 5, 2009, pp 1038) I get the following information
> for a particular dihedral
>
> dihedral V1
> V2 V3
> NA-CA-CS-CS -1.788 0.756
> -0.288
>
> When I use these values in the amber frc file,
>
> PK = V1/2 V2/2 V3/2
>
> PN = -3, -2, 1
>
> Is that right?
>
> Now I have the following queries:
>
> 1) does a negative value of Vn imply a phase shift of 180deg?How do I
> determine the phase shift?
>
Yes, the phase shift is 180 degrees. The Vn is the multiplicative constant
in front of a cosine term, and cos(x+180) = -cos(x)
2) Why are all the PK values in amber parm file positive and some in this
> format negative?
>
You can get the same effect by either setting the phase shift to 180
directly or by making the PK value negative. Most Amber force fields
always set PK positive, although I think some GLYCAM parameters use
negative PKs (and perhaps OPLS parameters do as well).
> 3) How do I determine the IDIV factor. Should I determine them by
> understanding how many torsions are there around that bond and divide PK by
> that value? Or do the authors already take that into consideration and give
> value for a single torsion (so in that case my IDIV factor will be one in
> all cases)
>
I guess I'm not understanding what data you have. If you are determining
the parameters yourself, you will need to take care of the dividing factor.
The typical convention adopted by Amber is to provide a non-unity dividing
factor only for generic dihedrals (e.g., X -CT-CT-X ). For specific
dihedrals (i.e., where all 4 atoms are specified directly and there are no
wild cards), we typically set the dividing factor to 1 and make sure that
the barrier height is properly adjusted for the superposition of the
expected number of dihedrals.
If you're taking the parameters from somewhere, you should assume that they
have either set IDIVF correctly or that IDIVF should be 1 and PK is
appropriately set (unless they specifically state otherwise).
If you are in doubt, it never hurts to validate.
HTH,
Jason
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Feb 21 2013 - 05:30:03 PST
