53 particles communicated to PME rank 1 are more than 2/3 times the cut-off out of the domain decomp

md_error.txt (10.5 KB)

I am trying to run a 100ns protein-DNA complex simulation using AMBERSB99 forcefield.

After equilibriation, this is the error that I got (md_error.txt).
53 particles communicated to PME rank 1 are more than 2/3 times the cut-off out of the domain decomposition cell of their charge group in dimension x. This usually means that your system is not well equilibrated.

How can I fix it? Please help out.

This usually means some large force is present that’s causing the simulation to blow up (see this page for more details and advice on how to fix it).

The exact cause usually depends on the simulation so more information is needed before figuring out what exactly is wrong (visualizing the system to figure out where clashes might be happening, plotting the potential energy during the minimization like in this tutorial, etc.)