Ewald for charged systems

From: Dongyue Liang (dliang25_at_uchicago.edu)
Date: Mon Oct 18 2021 - 12:55:27 CDT

Dear NAMD users,

I'm simulating a charged residue in solution with periodic boundary conditions. For my purpose in later research, I would prefer if no additional ions are added to the simulation box (for charge neutralization of the entire system), and I only care about the conformations. In other words, I won't mind if the energies calculated in the MD are wrong or differ by a constant, as far as the forces are correct.

I'm still hoping to take advantage of Ewald/PME for electrostatics. I found this old discussion in the mail list, and it appears that 1) the use of a uniform neutralizing plasma avoids the divergence problem of Ewald and gives correct dynamics; 2) with NAMD, people "implicitly" use the uniform neutralizing plasma method to treat Ewald for charged systems.


I tried to do MD on a charged system with PME with conventional commands (just like for a neutral system) with NAMD 2.14. The simulation runs "normally" but I'm not seeing additional output information regarding doing PME with a charged system. I did not find relevant information in the NAMD manual either. I'm wondering whether NAMD automatically "corrects" the Ewald with a uniform neutralizing plasma (or does some tweaks on the diverging term, etc.) if the system is charged. If not, is there a way that one may manually add it?

Any discussions or insights on this topic will be appreciated.

Best Regards,

This archive was generated by hypermail 2.1.6 : Fri Dec 31 2021 - 23:17:11 CST