VMD-L Mailing List

From: John Stone (johns_at_ks.uiuc.edu)
Date: Thu Aug 28 2003 - 15:46:52 CDT

Stephane, and others,
  Thanks for all of the the various Amber test data, we plan to use this to
update the Amber plugins to improve support for the variants of the
various files that it produces. While I was at Siggraph at the end
of last month, I had a brief opportunity to talk with David Case while
visiting Scripps to give a talk about recent work in VMD. David gave
me several good suggestions that may allow us to greatly improve how
the CRD reader works eliminating the silly need to choose "crd" or "crdbox"
depending on the type of file you load. I'm hoping that I'll have
time to address this before VMD 1.8.2 ships, or if not, very soon after
in a plugin-update to come a few weeks later.

Thanks,
  John Stone
  vmd_at_ks.uiuc.edu

On Thu, Aug 07, 2003 at 04:54:51PM +0200, Teletchéa Stéphane wrote:
> Le ven 01/08/2003 à 01:01, John Stone a écrit :
> > Hi guys,
> > If you guys can help me out by providing test files in the various
> > formats, I think I can put together an appropriate plugin for reading
> > all of these variations of files in short order. I talked to David Case
> > about some of this today while I was visiting Scripps, and he had some
> > good suggestions that I'm going to try when I return from Siggraph.
> > I'll email you guys about getting test data when I get back.
> >
> > Thanks,
> > John Stone
> > vmd_at_ks.uiuc.edu
>
> Here are my test cases (dna decamer in explicit solvent + counter ions)
>
> molecule.pdb
> molecule.parm
> molecule.inpcrd
> molecule_vmd.inpcrd
> molecule.rst
> molecule_vmd.rst
> molecule.mdcrd
>
> I've tried to be as explicit and exact as possible in the files :
>
> molecule.pdb : what it should look. (read as a pdb)
>
> molecule.parm : amber6 topology format. (read as a parm)
>
> molecule.inpcrd : input crd format know as inpcrd which is not read
> correctly in vmd. Please note that this file has no box coordinates (but
> it could in theory), so you must read it as a crd file. (read as a crd)
> If you read this file with crd only you'll see a drawing of the molecule
> completely messy with one line very far of the rest of the molecule, it
> seems that all n atom is read with coordinates n+1.
>
> molecule_vmd.inpcrd : here is the tricky file -remove the first line-
> and load it like the .inpcrd. Magically you'll see that the file is now
> correctly read. Be aware, that it is not correct but works (see
> discussion From Dr Caldwell and I).
>
> molecule.rst : restart file from amber run, output with the -r option.
> It contains coordinates for the molecule, velocities also and box info,
> read it with crdbox. The problem is the same as molecule.inpcrd when you
> read it, but there is another complexity coming from the velocities :
> the reader thinks it is a second frame although there is only one frame.
>
> Again, removing the first line (like molecule_vmd.inpcrd) solves the
> problem by 'tricking' it : molecule_vmd.rst
>
> Last, the correct molecule read correctly by the plugin :
> molecule.mdcrd (read it as crdbox) will display six frames (like this
> you can see how the transitions are managed).
>
> To be short, there are 3 differents types :
> - mdcrd which contains 10 columns low-precision coordinates, correctly
> read by vmd;
> - inpcrd which contains 6 columns high-precision coordinates, read
> correctly by tricking it;
> - rst which contains 6 columns high-precision coordinates plus
> velocities, read as two frames instead of one.
>
> To complicate it, each of the format can have a box info or not !
>
> All these files are real test cases coming from a real dynamics, and i
> hope they will help you solving the problem.
>
> I'm taking holidays tomorrow afternoon so if you have some questions or
> test you want me to make, please as before this date (i know it's hard
> ...) or i'll answer by the beginning of september (soon enough for 1.8.2
> ?).
>
> Stef
>
> PS : I've posted only to you as i'm afraid it will be logically thrown
> away on the vmd ML.
> --
> Teletchéa Stéphane <steletch_at_biomedicale.univ-paris5.fr> 

-- 
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