[Chimera-users] surface around combined "molecule"

[Elaine Meng]

On Jun 23, 2010, at 3:07 PM, Varela, Lawrence Rocco wrote:
> 
> ... Kind of a follow-up concern though is that Chimera is consistently crashing when I try to draw a cumulative surface around these molecules.  I can rewrite the question and scenario to the group email if you like, but here are the molecules I'm trying to draw the surface around if you would like to take a quick look.  The file with the word "combined" is obviously the one generated after saving and combining, the other is the original.  The reason why  I'm going through the process of first saving the molecules into a file with a single MOLECULE tag is because I want to draw a surface around the whole group, which I'm using to analyze some other results I have.  Am I doing this all wrong?

Hi Rocco,
In general to report bugs please use "Help... Report a Bug" in the Chimera menu, or if you really want to do it with e-mail, send it to chimera-bugs at cgl.ucsf.edu (the former is somewhat better because it automatically includes the Chimera version number and information about your computer system).

However, I'll just respond here since I think I understand what is happening.  It is known that the MSMS molecular surface code in Chimera is susceptible to numerical failures on some proportion of structures.  If you have a reasonably new version of Chimera (last year or so?), it should not crash but instead give an error message.  There are some workarounds, but since this is a numerical problem and not overtly deterministic (for example, my laptop shows the surface for that combined structure just fine, although it takes a while to calculate) it is not predictable as to which fix might work for a given structure, if any.  We have been working on our own code to replace that surfacing code for a while.  Here is one of the many earlier chimera-users messages about this, so I don't have to repeat myself:
<http://www.cgl.ucsf.edu/pipermail/chimera-users/2010-January/004713.html>

That said, it should be noted that I would expect a problem with your combined "molecule," which has lots of atoms practically on top of each other.  The code was designed to calculated surfaces of real molecules rather than envelopes of many superimposed molecules, so I would expect this to be even more likely to cause surface failures than a geometrically reasonable molecule.  

Here is another idea for showing the collective volume of a bunch of superimposed molecules: use the "molmap" command to make a density map from all those atoms, and then adjust the isosurface of the map to the desired contour level.
<http://www.cgl.ucsf.edu/chimera/docs/UsersGuide/midas/molmap.html>

For example, if your combined molecule is model #1, the command:

molmap #1 2.5 model #2

makes a density map based on Gaussians at the atom positions, where the 2.5 is "resolution" and proportional to the Gaussian width, and Gaussian amplitudes are proportional to the atomic numbers. You won't get exactly the same surface as the molecular surface since there is nothing that incorporates the atomic radii, but with your combined file, the command above, and moving the slider in Volume Viewer (automatically pops up when the command is executed and the map is created) to contour level of 0.1, the surfaces are remarkably close -- see figure attached below.  Even in cases where your superimposed structures don't give a MSMS failure, this is a much faster route to drawing an envelope around them!

I hope this helps,
Elaine
-----
Elaine C. Meng, Ph.D.                       
UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab
Department of Pharmaceutical Chemistry
University of California, San Francisco

the "molecular" surface is solid/multicolored, density map contour surface at 0.1 is white mesh:

-------------- next part --------------
A non-text attachment was scrubbed...
Name: grab.tiff
Type: image/tiff
Size: 133774 bytes
Desc: not available
URL: <http://plato.cgl.ucsf.edu/pipermail/chimera-users/attachments/20100623/e6a50408/attachment.tiff>