Re: [AMBER] Segmentation fault

From: Ben Roberts <>
Date: Thu, 29 Mar 2012 09:22:06 +1300

Hi Agostino,

On 29/03/2012, at 4:33 AM, Agostino Bruno wrote:

> Dear all,
> I am writing you because I tried to parametrize a ligand using xleap. I
> parametrized a ligand in a TIP3PBOX with 1 Cl- atom (so as to neutralize the
> net charge). During the parametrization stage everything seems work well. At
> the end I obtained the ligand.prmtop, ligand.inpcrd, and ligand.pdb files.
> First of all I checked the generated fiels using vmd, and everything seems ok.
> After that I tried to minimize the ligand using Amber11, I used one processor
> and the command line:
> sander -i -o minimization.out -p prova.prmtop -c prova.inpcrd
> -ref prova.inpcrd -r mini_restrt.
> But Amber almost immediately crashed with this error message:
> Segmentation fault.

Looking at the mdout file you attached, I see that the 1-4 VDW and 1-4 electrostatic (the latter shown as 1-4 EEL) energy terms are given as "NaN", as are the RMS gradient and the total energy. This likely means one of two things. Either two atoms in a 1-4 relationship are on top of each other -- in which case the denominators in the expressions for electrostatic and van der Waals terms go to zero -- or you have bogus parameters, typically a zero value where a non-zero value is expected. Since both electrostatic and van der Waals terms are affected, I'd suspect the former.

In short, check your structure for overlapping atoms.

Hope that helps,

AMBER mailing list
Received on Wed Mar 28 2012 - 13:30:03 PDT
Custom Search