[gmx-users] question about exploding

[Mark Abraham]

shensui830720 at aim.com wrote:
> I am a new user. I use gromacs to do MD. The system includes a small 
> protein and some water/ChCl3/Methanol solvents. I did it under nPT 
> (constant). However, it exploded after several minutes. And when I 
> looked at the trr file, I found the protein moved outside my box. How to 
> solve this kind of problem and what I need to do? Thank you very much!

Probably, you haven't relaxed bad geometric contacts before trying to 
simulate. A protocol along the lines of energy minimization, NVT 
equilibration, then NPT simulation is a good idea. Make sure you read 
the man pages for editconf, genbox, pdb2gmx and grompp so you understand 
the impact of all of the command-line options you are choosing. Trying a 
  simpler system first might help control your variables :-) Try to find 
a paper that describes simulation on your sort of system and copy their 
equilibration method.

Mark