AMBER Archive (2001)

Subject: Re: Amber, VMD and visualization

From: Warren DeLano (warren_at_sunesis-pharma.com)
Date: Thu May 24 2001 - 20:30:00 CDT

Hermann,

 One new alternative to VMD is PyMOL ( http://pymol.sourceforge.net ), which is
natively scriptable using Python (and redistributable open-source to boot).
The disadvantage is that you'd have to write out your coordinates (at least
temporarily) as numbered PDB files, and you'd have to compile
PyMOL+dependencies since there isn't an RPM yet.

 Assuming you had a directory full of numbered PDBs of the form
"prefixXXXX.pdb", you could load the trajectory into an object named "movie" by
running "pymol.com load.py", with the following 4 line program.

---BEGIN FILE "load.py"

from pymol import cmd
from glob import glob

list = glob("prefix*.pdb")

for file in list: cmd.load(file,"movie")

---END FILE

NOTE: This approach only makes sense to pursue if you are looking at hundreds
of states or less, with only a few thousand atoms per state. Otherwise
performance and memory considerations will become prohibitive. 

--
mailto:warren_at_sunesis-pharma.com
Warren L. DeLano, Ph.D.
Informatics Scientist
Sunesis Pharmaceuticals, Inc.
3696 Haven Ave., Suite C
Redwood City, CA 94063
(650)-562-3106 fax: (650)-556-8824

Hermann von Grafenstein wrote:

> Dear Amber experts,
> We are new to Amber. So far, we have managed to install Amber6 on Linux
> (Redhat 7). Now we would like to visualize (animate) energy minimization
> trajectories. As evident in the mail reflector, others have successfully
> used VMD for that purpose. What do we need to do, at the Amber end, for
> piping the coordinate outputs to VMD? And how do we build, at the VMD end,
> an interface that can accept output from Amber6? How can we get Amber to
> accept a connection request from VMD? How can we specify Amber to send the
> coordinates to a particular port? Does anyone have a Sander input file
> and/or a Tcl script for VMD?
> What about using gOpenmol? Has anybody used that?
> Please point us in the right direction. Any suggestions would be gratefully
> appreciated.
> Thank you very much.
> Hermann von Grafenstein and Harsh Parikh.