For a long time I am dealing with a problem with MD simulation.
I want to simulate a system composed of a lipid bilayer with two separated "baths" of water with ions inbalance. My protocol works in all my systems except one.
I have diagnosed that when I don't move any ions after equilibration, the system works (the simulation starts), but when I want to move one ion, the error occurs:
" Fatal error:
of the domain decomposition cell of their charge group in diamension x.
4 particles communicated to PME rank 2 are more then 2/3 times the cut-off out of the domain decomposition cell of their charge group in diamension x.
This usually means that your system is not well equilibrated."
The only files that GROMACS produce are: .log (with 0 steps), .xtc, .trr and .edr.
What would You suggest to do? I am dealing with the problem for a month and still no progress...
If you change coordinates, you have to re-minimize and re-equilibrate. If you attempt to continue using the previous state (including velocities), you are certain to introduce instabilities. Why are you changing ion coordinates?
I am trying to simulate electroporation by making the ions inbalance across the memebrane.
I have been doing the minimization before that, but now I will try to also equilibrate the system a little bit as You suggested.
Thank You, I will inform about the results!
Good to know about such a module! However, I don't want to simulate a double-bliayer system (because of computational costs) but I want to prepare the one with vacuum on both sides of the membrane. The method is described here ( Chapter Electrical Conductance of Lipid Pores
) if You are interested.- Badges
- Science topic