Hello Bala,
You are correct - traditionally dihedrals are considered independent with
respect to other internal coordinates when one does parameter optimization.
When you have a molecule that can form an intramolecular hydrogen bond, you
want to generate conformations, during a rotational scan of the dihedral,
such that the intramolecular hydrogen bond is disallowed. (Note, that I am
assuming that we are talking about an intrammolecular hydrogen bond that is
formed or broken during to the dihedral rotation, versus one that is remote
from the torsion of interest.) An example of this is in the optimization of
carbohydrate parameters (eg. Glycam06 and O-C5-C6-O5). The can often be
accomplished by putting the hydroxyl hydrogen into a trans conformation.
Sometimes you will also need to put a second constraint in your QM and MM
optimizations to prohibit the hydroxyl rotation. This can be conceptually
rationalized by considering what happens to a solute molecule in a hydrated
phase - most often the water will compete with the solute's intramolecular
forces to form an intermolecular hydrogen bond, and thus you underlying
solute potential energy needs to reflect that in some way.
So all of that are for Class I force fields, like Amber's. However, in
Class II force fields, you have cross terms that can explicitly account for
the coupling between internal coordinates. These can take the form of
bond-angle, angle-torsion, torsion-torsion, etc. terms. These extra terms
in the force-field equation add additional costs for calculating the
forces, but they are also tend to be more accurate. Allinger's MM4 program
is a good example of a Class II force field that achieves very high
accuracy.
I would suggest to do some addtional reading - a quick search gave me
this recent paper that is informative:
Vanommeslaeghe, K., Guvench, O., & MacKerell, A. D. (2014). Molecular
Mechanics. *Current Pharmaceutical Design*, *20*(20), 3281–3292.
Hope that helps,
Karl
Karl. N. Kirschner, Ph.D.
Research Associate
Bonn-Rhein-Sieg University of Applied Sciences
Grantham-Allee 20, 54757 Sankt Augustin, Germany
Twitter: .k_n_kirschner <https://twitter.com/k_n_kirschner>
On Thu, Jun 22, 2017 at 10:42 AM, Bala subramanian <
bala.biophysics.gmail.com> wrote:
> Dear Amber users,
>
> I am doing ff parameterization for a molecule (a amino acid mimic), where
> QM optimized energies of a dihedral scan vary depending on torsions of
> another dihedral, which facilitates intramolecular hbond.
>
> I assume that in ff parameteriztion each dihedral is considered independent
> of others while deriving the torsional term. I would like to know any
> caution/suggestion from field experts how correlated dihedrals can/should
> be treated to obtain the torsional terms.
>
> I apologize if this query is more general and not specifically about any
> Amber code.
>
> Thanks,
>
> Bala
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Jun 22 2017 - 04:00:03 PDT
