From: Bennion, Brian (
Date: Mon Dec 17 2012 - 10:19:02 CST

From: Axel Kohlmeyer []
Sent: Friday, December 14, 2012 11:07 PM
To: Bennion, Brian
Subject: Re: vmd-l: visualizing multiframe trajectory simultaneously with multimolecule

On Sat, Dec 15, 2012 at 12:06 AM, Bennion, Brian <> wrote:

>> Thank you for any suggestions
> Extensions -> Visualization -> Multiple Molecule Animation
> does this not work?

> NO, but you like more details than it doesn't work. The best I can tell you is I can't even test it. It locks up the vmd session and takes most of my cpu cycles until I kill -9 the vmd session.
> More details.
> mol new test.psf test.pdb test.dcd
> 6000 frames are loaded into one molecule.

here is a mistake. for multimol anim you need to load
each frame into a separate molecule. if you need to
copy the choice of representations and selections,
you can use the "Clone Representations" plugin.

How will this help solve the problem? Now I will have X number of molecules that correspond to X number of frames of a protein simulation. I still have to find out a way to overlay the subsequent X number of mesh graphics objects (that were loaded as molecules) to the exact protein molecule from which it was calculated.
> mol new dir1/000000.mesh
> ...
> mol new dir999/000999.mesh

and, of course you need to add the graphics
to each molecule (=frame). loading them into
separate frames may cause other viz problems

I am confused by the comments above. Are you saying that the meshes that have been calculated should be added to each of the protein frames/molecules using the vmd graphics command language instead of loading them as separate molecules? I will look into how to do this in a frame independent manner. Thank you for the suggestion.

> if the mol new does not find a mesh it loads a blank molecule.
> Then I click through Extensions -> Visualization -> Multiple Molecule Animation
> It pops up a blank window with the correct label "Multiple Molecule Animation"
> Then it just sits there. Maybe there are to many "graphics molecules"?

dunno. maybe too many graphical elements?

> Is the extension compatible with graphics objects that are called molecules.
> Perhaps this is the sticking point?

it should not, but as i mentioned before,
it makes a different assumption about what
to show and how that doesn't fit with how to
load your data into VMD. i wrote that code
about six years ago and didn't touch it since.

> Anyway I will keep testing with smaller and smaller sets of graphics molecules until I can actually use the extension.

why not start with just 3 frames for starters?
that is how i did the debugging of the code...

As a first pass I trust that code should work, then I start debugging.

> In the meantime, I don't think that "one" slider will advance both the frame number of the 1st molecule and display a single graphics object. Both the slider in the vmd main window and the slider in the multiple molecule extension will have to be coordinated somehow.

not needed if you adjust the way how you are
loading the coordinate data and graphics.


> Brian

Dr. Axel Kohlmeyer
International Centre for Theoretical Physics, Trieste. Italy.