VMD-L Mailing List

From: John Stone (johns_at_ks.uiuc.edu)
Date: Fri Jun 22 2007 - 14:38:40 CDT

Yes, the whitespace in a PDB file MUST BE EXACT, that's what I meant
when I said it was a column-based file format. Please see the PDB
file format specifications for more details:
  http://www.wwpdb.org/documentation/format23/sect9.html#ATOM

Note the specific columns that deliniate the fields within an ATOM record.

Cheers,
  John Stone
  vmd_at_ks.uiuc.edu
 
On Fri, Jun 22, 2007 at 12:38:42PM -0700, Arneh Babakhani wrote:
> Yes, you're right, thanks. So the amount of whitespace in a pdb file
> must be exact? I thought most readers can parse that out? That's a
> very important lesson! Thanks for your help,
>
> Arneh
>
> Giovanni Bellesia wrote:
> >looks like your pdb is not formatted correctly:
> >Try this:
> >
> >ATOM 1 H XXX 1 2.00 2.00 0.00 H
> >ATOM 2 H XXX 1 0.00 2.00 0.00 H
> >END
> >
> >it worked for me.
> >
> >You also might want to take a look at this:
> >
> >http://www2.cs.uh.edu/~chandler/chem/PDBSpec.html
> >(9. Coordinate Section)
> >
> >>Thanks, let me try to clarify.
> >>
> >>Assuming the pdb is formatted correctly, when I load it up and query
> >>the coordinates of say the first atom, I get this:
> >>>Main< (PracticeVMD) 67 % [atomselect top "index 0"] get {x y z}
> >>{2.0 0.0 0.0}
> >>
> >>But this does not match up with the coordinates in the actual pdb
> >>file, where the coordinates for the first atom are {2.0 2.0 0.0}, b/c
> >>of what you stated below (the viewing transformation is independent
> >>of the atom coordinates).
> >>
> >>My question is, how can I load up the pdb without such a viewing
> >>transformation? In other words, when I do 'get {x y z}', I want the
> >>result to be the exact coordinates found in the pdb file.
> >>
> >>Thanks,
> >>
> >>Arneh
> >>
> >>
> >>
> >>John Stone wrote:
> >>>Hi,
> >>> I'm not sure what you're asking here. VMD preserves the original
> >>>coordinates of the structures after they are loaded. The viewing
> >>>transformations applied for graphical scaling/rotation/translation
> >>>are completely independent from the atom coordinates. You can
> >>>query the atom coordinates using a selection to verify this for
> >>>yourself:
> >>> set sel [atomselect top all]
> >>> $sel get {x y z}
> >>>
> >>>If you're writing PDB files by hand, you need to make absolutely
> >>>sure that you've got the PDB columns lined up since it is a column
> >>>based format. If you write the coordinates into the wrong columns,
> >>>you'll get gibberish from VMD (and any other program....)
> >>>
> >>>Cheers,
> >>> John Stone
> >>> vmd_at_ks.uiuc.edu
> >>>
> >>>
> >>>On Fri, Jun 22, 2007 at 11:10:06AM -0700, Arneh Babakhani wrote:
> >>>
> >>>>Hello,
> >>>>
> >>>>I have a brief question about loading up pdb files in VMD. I
> >>>>noticed that when I load up a pdb file, one of the atoms is placed
> >>>>at the origin {0 0 0} and the rest are plotted around that. For
> >>>>instance, for the simple pdb file:
> >>>>ATOM 1 H XXX A 1 2.00 2.00 0.00 H
> >>>>ATOM 2 H XXX A 1 0.00 2.00 0.00 H
> >>>>END
> >>>>
> >>>>the fist hydrogen atom is placed at {0 0 2}, the second is placed
> >>>>at {0 0 0}. So the distance between the two atoms is correct, but
> >>>>they're coordinates are different from what actually appears in the
> >>>>pdb file.
> >>>>So my question is, how can I load up the pdb file such that the
> >>>>atoms are placed in the 'absolute' coordinates of file? And how
> >>>>can I do that from the TCL command line, is there an option to 'mol
> >>>>load pdb whatever.pdb' ?
> >>>>Thanks,
> >>>>
> >>>>Arneh
> >>>>
> >>>
> >>>
> > 

-- 
NIH Resource for Macromolecular Modeling and Bioinformatics
Beckman Institute for Advanced Science and Technology
University of Illinois, 405 N. Mathews Ave, Urbana, IL 61801
Email: johns_at_ks.uiuc.edu                 Phone: 217-244-3349
  WWW: http://www.ks.uiuc.edu/~johns/      Fax: 217-244-6078