[gmx-users] AMBER ff99SB port problems

Mark Abraham Mark.Abraham at anu.edu.au
Sat Dec 8 00:13:25 CET 2007

Shozeb Haider wrote:
> Hi,
> I was wondering if any has expertise in using Amber ff99SB port (Pande's 
> lab). I am currently using port for 3.3.1 version of gromacs.
> When I solvate my protein and run grompp to generate a tpr file, it 
> gives me the following error
> checking input for internal consistency...
> calling /lib/cpp -traditional...
> processing topology...
> Generated 2628 of the 2628 non-bonded parameter combinations
> Generating 1-4 interactions: fudge = 0.5
> Generated 2628 of the 2628 1-4 parameter combinations
> Cleaning up temporary file gromppXORFn5
> -------------------------------------------------------
> Program grompp, VERSION 3.3
> Source code file: toppush.c, line: 1108
> Fatal error:
> [ file "/usr/share/gromacs/top/spc.itp", line 41 ]:
>             Atom index (1) in settles out of bounds (1-0)
> -------------------------------------------------------
>  From what it seems that the program is unable to communicate with 
> spc.itp file. However that should not be the case since the program 
> routinely does that when I run using other gromacs force fields. I have 
> also explicitly included the /share/gromacs/top direcly as well

Nope, it's finding spc.itp just fine. Have a look at what's in there, 
and if necessary check out http://wiki.gromacs.org/index.php/cpp for 
some hints on how it's working. Each GROMACS forcefield #defines a 
variable, and since this AMBER port is non-standard, the standard SPC 
file doesn't deal with that possibility. So when parsing spc.itp, grompp 
doesn't see any atom definitions. The first thing it does see is the 
settles, which are out of bounds because there are zero atoms. To fix 
this, either use some ffamber_spc.itp if one exists, or make one.


More information about the gromacs.org_gmx-users mailing list