[Chimera-users] Slow dealing with pdb files

[Francesco Pietra]

I am dealing with the average structure (a protein complex embedded in a POCP
membrane and water solvated) derived with Amber's ptraj from a 1.5 ns MD.

Opening this pdb file in 1.2470 Chimera has become extremely slow. The file is
6.4MB. First, below the screen it is warned "Ignored bad PDB record found on
line #", for lines from 1 to 114154. This may take some 10 minutes.

The pdb records read as in the following examples:

For the lipid:
ATOM     20 2C21 POP     1      25.569  20.201  48.492  0.00  0.00

For the protein:
ATOM  10915  HB2 ALA   117      44.211  74.567  28.832  0.00  0.00

For water:
ATOM  22264  O   WAT  1771      25.558  39.417  16.580  0.00  0.00
ATOM  22265  H1  WAT  1771      25.582  39.432  16.549  0.00  0.00
ATOM  22266  H2  WAT  1771      25.569  39.482  16.611  0.00  0.00

After that, the warning message changes to "Computed secondary structure
assignments (see reply log)" which lasts for longer than 1 hour and 20 minutes.
During this time, "top" command shows that python is using 12% MEM and 99% CPU.
Then, the graphics appears, with the membrane-protein-complex not centered in
the water box.

During all that time, I had to avoid doing anything else with the GNOME
interface. A second terminal for "top" had to be opened before launching
Chimera on another terminal window. Otherwise, spurious graphics become
superimposed to the Chimera window.

I could then carry out rapid mapping of the protein residues around the
single-residue ligand (select protein & :ligandname z<#), which was what I
wanted to do.

The same events occurred with shorter trajectories. At ca 0.7ns the time taken
by Chimera to work out the average-structure pdb was about 15 minutes. Clearly,
there is an exponential trend.

I can in part compare these events on VMD: the pdb file from 1.5ns MD is opened
in VMD in less than 1 minute and the resulting membrane-protein-complex is
centered in the water box.

Is all that caused by the not-updated atom naming by Amber? Or could the
execution be accelerated by calling some C routines by python?

Chimera was run on a modest desktop: Athlon 1GHz, RAM 1GB, a poor main board
(product: K7S5A, vendor: ECS, version: 1.0), Debian Linux i386.

Thanks

francesco pietra


      ____________________________________________________________________________________
Be a better friend, newshound, and 
know-it-all with Yahoo! Mobile.  Try it now.  http://mobile.yahoo.com/;_ylt=Ahu06i62sR8HDtDypao8Wcj9tAcJ