VMD-L Mailing List

From: John Stone (johns_at_ks.uiuc.edu)
Date: Thu Mar 29 2007 - 11:03:17 CDT

Hi,
  Are the CP2K DCDs being corrupted only under certain conditions
(e.g. if the program terminates abnormally) or is this a consistent
problem? You could write a Tcl script in VMD to reconstruct the
damaged DCD header if you knew what attributes it had before it
got corrupted, e.g. the atom count, whether there were fixed atoms,
which DCD format variant it was, file endianism, etc.

Most likely the time-economical solution would be to work with the CP2K
authors and figure out why the header is getting corrupted, and help them
get their DCD writing code fixed. The CP2K authors might want to look
at NAMD to see how it writes its DCD files. NAMD tries very hard to
deal with supercomputers that have flaky I/O subsystems, and recover
gracefully from various error situations as best it can.

  John Stone
  vmd_at_ks.uiuc.edu

On Thu, Mar 29, 2007 at 11:21:31AM -0400, Nichols A. Romero wrote:
> Hi,
>
> I saw similar topics discussed on this forum but mine may be slightly
> different.
>
> I am doing ab initio MD using CP2K.
> http://cp2k.berlios.de/
>
> Occasionally, it seems like the DCD files that are created don't write the
> header
> information (or part of it). This seems related to restarts, but I am not
> 100% sure,
> but this occasionally does happen.
>
> In any case, given this 'damaged' DCD file is there anyway to figure out
> what part of the
> header is missing and then add the missing information.
>
> I have that matdcd m-files tools if that helps.
>
> P.S. I do not believe this is related to a different DCD version, as I have
> been able to
> read DCD files which are compliant with the most recent DCD standard.
>
>
>
> --
> Nichols A. Romero, Ph.D.
> 1613 Denise Dr. Apt. D
> Forest Hill, MD 21050
> 443-567-8328 (C)
> 410-306-0709 (O) 

-- 
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