AMBER Archive (2001)

Subject: Re: Problem minimizing on SGI (SHOULD NOT BE HERE).

From: Andrew Hoofnagle (hoofnagl_at_SDSC.EDU)
Date: Thu Sep 13 2001 - 15:39:29 CDT

Thank you all for efforts in sorting out the problem.

Mike Crowley wrote:

That message indicates two things:
1) thereis something wrongwith your input files
2) Amber code has failed and made to a place in
   the program that theoretically cannot happen: there
   is no logical way to get that that print statement
   and thus the message "should not be here"

Do your input files work in any other way with sander or sander_classic?
For starters, make sure that your input files are for periodic systems,
sander will not work with input files for non-periodic systems
with igb=0.

Check the end of both the inpcrd and prmtop files for
box size and angle information. Also check to make sure
that the number of coords in inpcrd matches the number
of atoms. Finally make sure that your choice of NTX matches
what you have for an inpcrd file.

The real problem turned out to be that I had too many atoms (>100,000).

Andy