[Chimera-users] Smoothing a trajectory

Tom Goddard goddard at sonic.net
Tue Sep 20 12:01:43 PDT 2011

Hi Dan,

   Chimera molecule models take lots of memory, about 2 to 4 Kbytes per 
atom.  If your 200 frame models had 5000 atoms each that would be one 
million atoms taking about 2-4 Gbytes.  You would certainly run out of 
memory and likely crash with a 32-bit version of Chimera (addressible 
memory of 4 Gbytes).  A 64-bit version of Chimera might get further if 
your machine has more 8 Gbytes of actual memory.  But you can see it 
won't scale to thousands of models.

   So I would suggest opening a small batch of frames (50) as individual 
PDB models, compute and play the morph, then close that batch of models 
and open the next (50) and compute and play the morph.  All the while 
you record the movie.  The pauses during the morph computations and 
model opening/closing will not appear in the movie because the Chimera 
graphics is not being redrawn at those times.  This method is a bit more 
complex though.  Maybe it would be slightly simpler to just morph 
between two frames a time.


> Hi everyone,
> For the purpose of making educational content, I'm interested in 
> smoothing out the often jumpy transitions from frame to frame in a MD 
> trajectory. My approach was to write two python scripts (with critical 
> help from Eric of course); one goes through the trajectory and saves 
> pdbs of each desired frame, while the second opens these pdbs and 
> creates a morph, interpolating from one pdb to the next.
> It seemed like I was almost there, and it looked great when I tried 50 
> frames.....but Chimera crashed when I tried morphing 200 frames, 
> adding 3 frames of interpolation between each.
> Is there a better way to add transitions in between (ideally thousands 
> of) frames of an MD trajectory?
> Thanks
> Dan
> -- 
> ____________________________
> Daniel Gurnon, Ph. D.
> Associate Professor of Chemistry and Biochemistry
> DePauw University
> Greencastle, IN 46135
> p: 765-658-6279
> e: danielgurnon at depauw.edu <mailto:danielgurnon at depauw.edu>
> _______________________________________________
> Chimera-users mailing list
> Chimera-users at cgl.ucsf.edu
> http://plato.cgl.ucsf.edu/mailman/listinfo/chimera-users

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://plato.cgl.ucsf.edu/pipermail/chimera-users/attachments/20110920/60fa0b54/attachment.html>

More information about the Chimera-users mailing list