Calculating RMSE of PME forces

GROMACS version: gromacs-2018.6-gpu-mpi
GROMACS modification: No

We are studying the dielectrics of solvated proteins and I recently read a paper that points out that the PME parameters greatly influence the dielectric constant of water. Upon further reading I found that, regardless of system size, the dimensionless RMS error of the PME electrostatic forces should be on the order of 10 ^-4 to obtain an appropriate estimate of the dielectric constant.

So I wanted to ask, if there are any programs within GROMACS or third party packages where I can calculate the RMS of the PME forces to be sure that I am using appropriate order and grid spacing? I haven’t been able to find anything so far.

Related Citations:
J. Chem. Phys. 128, 214106 .2008
J. Chem. Phys. 132, 244109 .2010.

Thank You

The decisive factor is fourierspacing/rcoulomb ; this should be < 0.125. With larger rcoulomb you can afford larger fourier-spacing for same accuracy. Other people in this forum might have deeper insight into the full derivation of this.