[Chimera-users] PDB

Dougherty, Matthew T matthewd at bcm.tmc.edu
Wed Jun 23 14:07:53 PDT 2010

Hi Eric,

Yes this is closer to the situation, lots of monomers making up a complete virus.  But the issue is not data/memory size, but rather provenance & transformations.

The researchers are constantly tweaking the PDB data prior to submission.  I might get a dozen intermediate versions (eg, 12x8 PDB structures) before doing a final animation on the final PDB data sets.  They tend to have the same number of residues/atoms, although there may be a few additions of side chains or slight changes in backbone xyz coordinates.

Meantime, I am evolving these session files, with different savepos camera keyframes, so updating the PDB files becomes more convoluted as the session file becomes more complicated (e.g., introduction of corresponding EM densities, use of dozens of keyframes), plus the entanglement of the camera transformations and the model transformations.

From: Eric Pettersen [pett at cgl.ucsf.edu]
Sent: Wednesday, June 23, 2010 3:51 PM
To: Dougherty, Matthew T
Cc: chimera-users at cgl.ucsf.edu BB
Subject: Re: [Chimera-users] PDB

Perhaps the situation is that you have the same session using dozens or hundreds of copies of the same PDB file?


On Jun 23, 2010, at 1:47 PM, Eric Pettersen wrote:

On Jun 23, 2010, at 12:48 PM, Dougherty, Matthew T wrote:


When PDF files are read by Chimera and the session is written, the PDB data becomes part and parcel with the session files.  That is, Chimera does not go back to the PDB files when it reloads the session, it uses the PDB data within the session file instead; as contrasted to density volume files, which are referenced by the session file, and the density volume files reloaded when the session file is read by Chimera.

Is there a way to force the session file not to save the PDB internally to the session file, and instead reference and interact with it similar to the way volume files are managed?

Hi Matthew,
Briefly: no.  For a variety of reasons:

1) Structural data didn't necessarily come from PDB files.
2) What to do about structure deletions and additions?
3) Session is less transportable/publishable (also occurs with volume files separate from sessions)

I don't see any upside really.  You wouldn't be able to swap in a different PDB file because there would structure attributes stored in the session file (charges, conservation scores, etc.) that would need to be mapped on a 1-for-1 basis with atoms/residues of the original PDB file.  I guess if your PDB file was huge there would be some space savings from having multiple sessions referencing the same PDB file.  Nonetheless, the space issue is a lot less important for structure files than volume files.


                        Eric Pettersen
                        UCSF Computer Graphics Lab

Chimera-users mailing list
Chimera-users at cgl.ucsf.edu<mailto:Chimera-users at cgl.ucsf.edu>

More information about the Chimera-users mailing list