AMBER Archive (2001)

Subject: Re: coordinate blowup in sander

From: jim caldwell (caldwell_at_heimdal.compchem.ucsf.edu)
Date: Fri Jun 15 2001 - 12:54:53 CDT

The only thing you can do is to rerun the calculation,
this time specify that the coordinate file is saved
in binary format. However, the presence of "******"
in the file probably means something else is going
wrong as well.

jim

On Fri, 15 Jun 2001, Michael Cooney wrote:

> Dear Amber users,
>
> In carrying out an MD simulation beyond 1 nsec (1000 psec), I ran into a
> crippling problem in the mdrestart file, namely that some coordinates
> "blew up" and registered as ******** (fortran for overflow). The result is
> that sander can't use the restart file for further calculations. The job
> was running on a campus machine and exceeded its cpu time limit, but I
> don't think this is the real cause of the overflow---the rst file was
> written before the job terminated. The unexpected character Fortran
> warning occurs when trying to restart the md run. If needed, I can send a
> few sample lines from the rst file.
>
> Any help would be greatly appreciated.
>
> Thank you,
>
> Michael Cooney
> Dept. of Biochemistry and Molecular Biology
> The University of Georgia
> Athens, GA 30602, USA
>
>
>

----------------------------------------------------------------------------
James W. Caldwell (voice) 415-476-8603
Department of Pharmaceutical Chemistry (fax) 415-502-1411
Mail Stop 0446 (email) caldwell_at_heimdal.ucsf.edu
513 Parnassus Avenue
University of California
San Francisco, CA 94143-0446
----------------------------------------------------------------------------