VMD-L Mailing List

From: Axel Kohlmeyer (akohlmey_at_gmail.com)
Date: Mon Apr 05 2010 - 16:05:34 CDT

On Mon, 2010-04-05 at 17:47 -0300, Ranyere Deyler wrote:

> Hello again, and thanks again for your help. Well, it did work now,
> the "topo retypebonds". I don't know why it didn't work before,
> must have done something wrong. But, it's good to know that it can be

PEBCAC is the most common reason for stuff that "doesn't work". ;-)

> done like that. I would spend lots of time doing it manually. So,
> Axel, you were saying, is it possible to guess the angles and
> dihedrals using "topotool" for the CNT? And them write down the psf

i have never tried this on a CNT. as i wrote before, i have used
AIREBO or Tersoff potentials for them instead. but yes, this should
be possible. what exists in topotools are functions that can derive
angle and dihedral definitions from the bond topology. for "flat" bonds,
it can also guess the impropers, but this has been tested on CNTs.

i would thus suggest to first construct a _very_ short nanotube
and compare the bonds, angles, dihedrals and impropers with what
you would expect from the force field that you want to use. so
what you get is some kind of educated guess. in many cases this
is just fine and can take away the burden of writing a complex
script or even do things manually (with even more potential for error).

keep in mind that what topotools does is only generate the
definitions, finding the right choice of parameters is up to you.
psfgen works differently by comparing your residue and atom names
with a residue database and has in this residue database all
the required settings and definitions.

> file the same way? Do I need to enter with some parameter? Or it does
> it automatically? Because, when i'll be needing the CNT to be not
> fixed i would need those in the psf file as well.

i suggest to first very carefully study the literature.
there are different approaches to this issue and they all
have advantages and disadvantages. as with all classical MD
simulations you always run the risk of getting useless results
from using an inadequate model.

cheers,
   axel.

> Thanks a lot. 

-- 
Dr. Axel Kohlmeyer  akohlmey_at_gmail.com
http://sites.google.com/site/akohlmey/
Institute for Computational Molecular Science
Temple University, Philadelphia PA, USA.