Generating novel structure

From: Fangqiang Zhu (
Date: Tue Mar 02 2004 - 13:38:12 CST


Recently quite a few users asked about generating structures for
some untraditional molecules, e.g., carbon nanotubes. Here I just
wanted to share some of my experience with you.

Usually the coordinates of these types of molecules can be
expressed in some mathematical forms, and therefore no
"crystal structure" is needed. However, if you want to use
psfgen, you need to write a topology file by yourself, because
traditional force fields designed for biomolecules certainly
don't provide topology for these exotic things. For more details
about topology files, see

If your molecule is big and has regular pattern, it may not be a
good idea to write the topology file by hand; instead you can use,
say, some scripting language to write it. Also, for any molecule,
there could be many ways to generate its structure, and there's
not really a "standard" method.

As an example, here's a Tcl script I wrote a while ago to
generate the psf and pdb files for armchair type of carbon
nanotubes of any size:
This script first writes a topology file and a psfgen input file,
and then executes psfgen to generate the structure. I used this
script to generate the structures for the nanotube tutorial in the
TCB summer school as well as for our earlier paper.

Hope some of you find it helpful. I'm sure many of you can do
the same thing with Perl, shell language, etc., possibly in an
even simpler way.

Good luck,

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:37:23 CST