AMBER Archive (2009)
Subject: Re: [AMBER] average torsional energy
From: Francesco Pietra (chiendarret_at_gmail.com)
Date: Mon Jun 08 2009 - 04:11:52 CDT
Suppose one wants to see how the dihedral term
Summation Vn/2 [ 1 + cos(n omega - gamma)]
and the 1-4 interactions (if they are the most relevant ones to my
original question) are treated on the code, is it possible to have an
indication which section of the relevant *.f files should be studied?
To understand the physical correspondence to Vn as it is treated
Suppose further one wants to intervene on that Vn (intervene means in
addition to what the code is now doing), are there relevant data from
the simulation output which may serve as the basis, if not the
empirical term DIHED? Or such an oversimplified route is only destined
to wasting time?
On Mon, Jun 8, 2009 at 10:29 AM, Karl
> Let's make it a $0.07 worth. From a force field development point of view the
> dihedral terms are usually created last, after bond, angle, L-J terms, and
> partial atomic charges. As such, and in my opinion, their roles is to account
> for any left over unaccounted quantum effects (e.g. anomeric effect, guache
> effect, e- conjugation, etc). As Adrian said, much of the energy seen in
> torsion rotation comes from the nonbonded terms, but any through-bond quantum
> effects is captured by the torsion terms. I would be careful to say that
> torsion terms are just a fudge factor, because they do model some chemistry
> in the same way that a bond term models stretching. Never-the-less, they also
> are used to to correct the rotational PE curve for the physics that has not
> been accounted for in the bond, angle, and nonbonded terms. They are also
> often adjusted in order to obtain better agreement to an experimental
> observable in an MD simulation (eg reproducing secondary elements in
> proteins), making their use as a fudge factor even more.
> The interpretation of a V1, V2, or V3 term is useful and is based on the
> orbital hybridization model (eg the rotation about the central bond in a
> sp3-sp2=sp2-sp3 torsion is represented by a V2 term). The differences in how
> one views each force field term, and how one approaches their
> parameterization is the reason why there is an art to parameterization :) .
> Karl N. Kirschner, Ph.D.
> Fraunhofer-Institute for Algorithms
> and Scientific Computing - SCAI
> Department of Simulation Engineering
> Schloss Birlinghoven
> 53754 Sankt Augustin, Germany
> Tel: +49 (0) 2241-14-2052
> Fax: +49 (0) 2241-14-1328
> On Sunday 07 June 2009 18:14, Carlos Simmerling wrote:
>> I add my 2 cents in support of Adrian's 3- clearly the DIHED energy is not
>> the energy for rotating around a bond. I also see this misinterpreted in
>> the literature quite often. As Adrian says, the DIHED is an empirical
>> correction term in most cases.
>> On Sun, Jun 7, 2009 at 12:09 PM, Adrian Roitberg
>> > Francesco and all.
>> > I am not fully sure I understand your question, but I will give a more
>> > general answer.
>> > My very strong opinion, shared by most practitioners, is that the DIHED
>> > term (torsions) can never be interpreted in ANY way.
>> > I also personally believe that one should be very careful in interpreting
>> > the other terms also (vdw, elec,etc) but that is a more complex
>> > discussion. Why is DIHED particularly bad? because it is a clear fudge
>> > term. While one can assign a physical interpretation to the other terms
>> > (after all, we have the 'idea' of a bond in our heads), the torsional
>> > potential could be thought as if if comes from the electrostatics and vdw
>> > terms only. The torsional term itself is simply a correction to get the
>> > force field to match a QM torsional potential.
>> > In short, the torsional term is NOT the torsional potential ! The
>> > torsional potential, from where minima and barriers are extracted, comes
>> > from the total energy vs torsional angle: a scan.
>> > This usual confusion has some troublesome uses in the literature. Some
>> > people will set the torsional term to ZERO, and call the new potential
>> > one with 'reduced barriers'. This is non-sense. Setting a torsional term
>> > to ZERO could do a number of things including: changing the location of
>> > the minima, changing the number of minima, increasing the barriers,
>> > decreasing the barriers, etc etc.
>> > Just my three cents' worth of opninion.
>> > Adrian
>> > Francesco Pietra wrote:
>> >> My I ask about the practice of correctly evaluating from the DIHED
>> >> term (with Amber 9 or 10) the average torsional energy for an alpha
>> >> helical stretch where a major bending at a GLY residue is observed?
>> >> thanks
>> >> francesco pietra
>> >> _______________________________________________
>> >> AMBER mailing list
>> >> AMBER_at_ambermd.org
>> >> http://lists.ambermd.org/mailman/listinfo/amber
>> > --
>> > Dr. Adrian E. Roitberg
>> > Associate Professor
>> > Quantum Theory Project
>> > Department of Chemistry
>> > Senior Editor. Journal of Physical Chemistry
>> > American Chemical Society
>> > University of Florida PHONE 352 392-6972
>> > P.O. Box 118435 FAX 352 392-8722
>> > Gainesville, FL 32611-8435 Email adrian_at_qtp.ufl.edu
> AMBER mailing list
AMBER mailing list