[gmx-users] QM guy at wits' end

Justin Lemkul jalemkul at vt.edu
Wed Nov 20 20:23:11 CET 2013


On Wed, Nov 20, 2013 at 2:07 PM, SAllen <allense at email.unc.edu> wrote:

> Hello,
>
> I want to run MD simulations on a series of macrocycles that contain a lot
> of non-standard residues (dehydroalanines, dehydrobutyrines, 1,3-thiazoles,
> and a pyridine for good measure). I've been coming the internets for a
> reasonable guide on how to develop parameter and topology files for my
> structures, and I'll I've found are sites saying that it's too hard. Does
> anyone have any good guides for developing (reasonable!) topology and
> parameter files? I have access to lots of QM programs (gaussian, spartan,
> even schrodinger).
>
>
Parametrization generally is too hard for anyone completely new to
computational chemistry, but if you've got experience with QM, it's
probably quite easy.  The main issue here is that there truly is not one
answer to this question.  One must first decide on a suitable parent force
field under which you can parametrize your molecules.  If you're deriving a
lot of things from scratch, almost anything can be made to work, so you
choose the force field whose methods are easiest to follow or the one(s)
for which a generalized force field for small molecules is available.
 AMBER and CHARMM are great choices here, because they both have
generalized force fields for arbitrary molecules (GAFF and CGenFF,
respectively).  The parametrization methods for these force fields are very
well detailed in the literature.  That's what you need to follow;
parametrize the molecule in the same way that other model compounds were
parametrized so that the underlying model is internally self-consistent.

CGenFF even has a web server (ParamChem) that will generate pretty good
starting topologies for just about anything you provide for it.  You can
then refine the parameters according to the 2010 paper (which is really a
case study on pyrrolidine and works through the protocol in very nice
detail).

-Justin

-- 

==========================================

Justin A. Lemkul, Ph.D.
Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 601
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441


==========================================


More information about the gromacs.org_gmx-users mailing list