[gmx-users] Rectangular vs triclinic boxes

[Erik Lindahl]

Ivano Eberini wrote:

> Hi,
> we are performing some MD of lipocalins with/without specific ligands.
> We used rectangular boxes for all simulations (3 ns / each). 3 MDs
> finished OK, but the fourth one always stops with this fatal error:
>
> "Warning: Only triclinic boxes with the first vector parallel to the
> x-axis and the second vector in the xy-plane are supported.
>           Box (3x3):
>              Box[    0]={         NaN,          NaN,          NaN}
>              Box[    1]={         NaN,          NaN,          NaN}
>              Box[    2]={         NaN,          NaN,          NaN}
>           Can not fix pbc.
> Fatal error: ci = -1 should be in 0 .. -1 [FILE nsgrid.c, LINE 214]"
>
> We tryed with different PCs: PIII 733 and G4 altivec enabled, with the
> same result.
> Someone knows what happened and, if a solution exists, how to solve this
> problem?
> Thanks in advance and happy new year.

Hi,

It seems as if both I and David are away from work these days, so I
apologize if our
responses are a little sporadic right now...

Anyway - is there something special with the 4th simulation, and are you
able to run
any steps at all? From the above error my first guess would be that
something happens
to your system so you get a very high veolcity somewhere (e.g. close
contact), and
then everything crashes (since the NaNs are propagated to the box size).
Have you
tried a smaller timestep and/or more minimization?

Cheers,

Erik