[gmx-users] Simulation not running, system exploding.

[Alan Dodd]

Have you checked the structure with ngmx?  As it uses
the gromacs running topology as input, it can often
show up gross errors in your system.  I had a system
explode when I started, which turned out to be due to
all atoms somehow being reduced to a 1D smear along
the Z-axis.

--- Fernando Mattio <mattiofer at gmail.com> wrote:

>    Hi dear Gromacs users,
> 
>    I need some help in my simulation. I am trying to
> put in a cubic box a
> protein next to an ice plane and then i solvate the
> system and try to run
> the simulation. As the pdb2gmx doesnt´ work with
> water molecules, I
> proceeded as the following:
>    First  I used  the  pdb2gmx  only  with the
> protein  to get a  topology
> file. Then  I tried to create an .itp file of the 
> ice molecules, and for
> that I used the water model spc.itp (but I changed
> the nomenclature in the
> file, everything that was called SOL I changed and
> called OHH).
>    Later I included this .itp file in the topology
> file, paying attention to
> the number of molecules. After that  I  used
> editconf to create a box and
> genbox to  solvate  the system (From the .gro file
> with the protein next to
> the ice plane). By this way in the topology was
> included the water molecules
> (called SOL) by the .itp file of the water. Then in
> the parameter file I
> tried to establish differences between the ice and
> water molecules by
> applying frozen groups options to the OHH molecules.
>    However the funny thing is that when I first used
> part of the ice plane
> (with only 26 molecules that were somehow far away
> from the protein) the
> simulation  finished, but when I  used the whole
> plane (round 1200
> molecules) the simulation crashes in no more than 5
> or 10 steps.
> 
>    The message is always like this :
>    Fatal error: Number of grid cells is zero.
> Probably the system and box
> collapsed.
> 
>    And there are warnings like:
>    1-4 interactions between 25 and 33 at distance
> 1,322 which is larger than
> the 1-4 table size 1 nm, This usually means your
> system is exploding.
> 
> 
>     I tried to increase the table size but as I
> increased this value the
> distance always increased together, and I got agai
> the same warning with
> different values,  so this didn´t help. I tried also
> to perform energy
> minimization before solvating the system, but
> something bizarre happened,
> part of the ice plane just changed position with the
> protein... like the
> visualization was the same, but water molecules were
> in the position where
> the protein was and the protein was in the position
> where  these water
> molecules were before.
> 
>    I tried then to put he whole protein a little bit
> far away from the plane
> to see what happens, and when I runned the
> simulation (without energy min.)
> the same birarre thing happened (part of the plane
> changed place with the
> protein).
> 
>    I am still a begginer in Gromacs, and I really
> don´t know what could be
> the reason for these errors... It is also the first
> time that I am dealing
> with molecular dynamics simulation. Someone wrote me
> that it could be
> something related to topologies, but i really
> couldn´t find where the error
> was. So I ask for your kindness help. Suggestions
> and advices will be
> welcome,
> 
>    Best regards,
>    Fernando Mattio
> > _______________________________________________
> gmx-users mailing list
> gmx-users at gromacs.org
> http://www.gromacs.org/mailman/listinfo/gmx-users
> Please don't post (un)subscribe requests to the
> list. Use the 
> www interface or send it to
gmx-users-request at gromacs.org.


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com