[gmx-users] Simulation Break => Due to bad domain decomposition ? (MPI) => Gromacs Version 3.3.3

[Mark Abraham]

Chih-Ying Lin wrote:
>  
>  
> Hi
> I used Gromacs version 3.3.3.

OK, but unless you need scientific continuity with earlier work, use the 
much faster 4.0.7. Additionally, domain decomposition was not 
implemented before 4.0, so either you're mistaken about your subject 
line, or you're using a 4.0.x version of mdrun, or something.

> My simulation system = one protein + 20 ligand + water molecules ( 7x 7x 
> 7 ) 

What does (7x 7x 7) mean?

> MPI setting => #PBS -l nodes=4:ppn=4,arch=x86_64   => 16 nodes in total
> After doing the energy minimization, => the potential energy is 
> extremely high ( say, ten to the 9th order )
> I visualized the " Simulation-System-EM-solvated.gro " after the energy 
> minimization. 
> Then, I found that the Simulation Sysmtem is devided into 16 
> domains very clearly and the molecules (protein, ligand, and 
> water) break into atoms in the boundaries.

I can't reconcile your description with any known behaviour of any 
version of GROMACS. Can you post "before" and "after" screenshots on a 
webpage please (e.g. photobucket).

The odds are excellent that a newish user is doing something wrong, 
rather than there being a novel problem with an old version GROMACS, 
however.

> I have checked that the 20 ligands are not overlapped each other and are 
> not overlapped with protein from the beginning.
>  
>  
>  
> More, i have created 10 alike system and each is with "one protein + 20 
> ligand + water molecules"
> Two of them get the "minus potential energy" after energy minimization 
> and I can continue running the MD simulation successfully.
> Others of them get the "extreme high positive potential energy" and the 
> system is devided into 16 domains after energy minimization and the 
> simulation broke afterall.

Well, some of them may be OK, and some are ill-formed. This is not 
evidence of a problem with GROMACS, but might be evidence of an 
ill-formed workflow.

> With one protein + 10 ligand + water molecules, ( 6 x 6 x 6 )
> There is no problems like that.

You got lucky, or used a different procedure, or something. Without a 
full description of your preparation protocol, there's nothing to be said.

Mark