Notepad++ will do a great job of editing the .py files. I've found that this is an extremely convenient way to accomplish certain tasks, such as precise alignment of multiple clipping planes. However, you definitely have to know what you're looking for when digging through the files. On a related issue, I have a question regarding the saved sessions. In some of the viral capsid models I've created a saved session is enormous (about 60 megabytes) and fails to compile either automatically when loading the .py session or manually using the python.exe included with Chimera. It seems to be a result of excessive explicit attribute assignments that I don't really care about (color etc) since they are easy enough to set once the model has loaded. Does anyone have any suggestions or experience with this issue? Jonathan PS: Check out the most recent issue of Science. Although the editors chose the least visually impressive submission, non of the variations would have been possible without Chimera: great program.
As per Elaine's mail, great cover! As for the session issue, since the last production release I've completely redone session handling. There are no longer PDB files embedded in sessions. Instead, structures are built directly by the session. Also, attributes that are the same for all atoms (or residues/molecules) are recorded once instead of once per atom. As a result, sessions are smaller, use less memory when restoring, and restore faster. This new session code is in the 1.2224 snapshot release. If you try out 1.2224 and you still can't get a newly-saved session to restore (i.e. a session that was written using 1.2224), then let me know and I'll investigate to see if there are further bottlenecks I can eliminate. --Eric Eric Pettersen UCSF Computer Graphics Lab pett@cgl.ucsf.edu http://www.cgl.ucsf.edu On May 22, 2006, at 4:17 PM, Jonathan Hilmer wrote:
Hi Jonathan, I would guess your session files for virus capsids are large due to having alot of atoms because you have loaded many copies of the capsid protein molecule. If this is the case than Eric's suggestion to use the latest Chimera snapshot (1.2224) is your best bet. The only other options being to save fewer atoms, or use a machine with more memory. You can create the *.pyc file on a different machine with more memory that need not have interactive graphics and transfer it back to your visualization machine if necessary. The *.pyc file is platform independent (windows, mac, linux, sgi, ...). Another possible cause of large session files is if you are looking at virus models with a large number of chains. For instance I made a model of a crystal lattice of 50 copies of a small virus (satellite tobacco mosaic virus, 1a34) with asymmetric unit having 6 chains. That gave (50 capsids) x (60 asym units per capsid) x (6 chains per asym unit) = 18000 chains. The session file was about 19 Mb. There were only 3500 atoms loaded (just one asym unit) and the space in the session file for that was negligible. So the session file chews up about 1 Mbyte per 1000 multiscale chains. Unless you are making some extremely large molecular assemblies your sessions are probably mostly filled with atom data, not multiscale chain data. Thanks for making the Science cover image! I am interested in making virus architecture movies in Chimera. Here is an example using my current favorite capsid, orthoreovirus PDB 2cse: http://www.cgl.ucsf.edu/chimera/animations/movies/reovirus.mov It is in the Chimera animation gallery http://www.cgl.ucsf.edu/chimera/animations/animations.html Tom
participants (4)
-
Eric Pettersen -
Jonathan Hilmer -
S Joshua Swamidass -
Thomas Goddard