pairInteraction and PME

From: Jason Swails (
Date: Wed Apr 03 2013 - 15:58:21 CDT

Hello everyone,

I have a question about how pairwise electrostatic interaction energies are
calculated when using PME. What I want to do is to calculate the energy of
a single water molecule with every other atom in the system (as a way of
approximating a one-step FEP free energy in which I annihilate the water).
 I can set up the calculation just fine and I even get back reasonable

What I can't figure out, though, is where the number comes from. I
understand that the reciprocal space sum is not pairwise decomposable (and
Axel says as much here:
 Nevertheless, NAMD gives me reasonable numbers. I can envision this
calculation being done in one of two ways (admittedly not much thought has
been put into alternatives):

1) Forget about the reciprocal space sum, just decompose the direct space
sum. Use a large direct-space cutoff and a diffuse neutralizing charge
distribution and call it good enough.

2) Perform 2 Ewald calculations: one with the whole system and the other
with the single water molecule given charges of zero, then take that
difference... or something along those lines. (Maybe you need a 3rd, where
the surroundings are given a charge of 0 and the water has a charge?)

I'm guessing the approach actually taken is (1), but I haven't been able to
find any documentation or discussions on the mailing list actually stating

Any help is appreciated.


Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate

This archive was generated by hypermail 2.1.6 : Tue Dec 31 2013 - 23:23:07 CST