Re: how does NAMD calculate intramolecular electrostatic interactions between Drude pairs?

From: Ana Celia Vila Verde (
Date: Mon Nov 23 2015 - 09:27:00 CST

Hi everyone,

I've actually found the answer to my original question, so I post it now
in the hope that it will be helpful to others.

Consider the artificial ion OS1-S-OS2, (i.e., half a sulfate to make
things easier). S is the sulfate, OS1 and OS2 are the oxygens; the
oxygens are polarizable but S is not.

Since S is not polarizable, it has only one associated, fixed, charge,
called core charge. In contrast, you can think of the total charge in
each oxygen as the sum of three charges: a core charge, qi, equal to the
nominal partial charge on that atom, and the Drude charges, qD and -qD,
associated with the electroneutral oscillators.

In sulfate, 1-2 and 1-3 electrostatic interactions between core charges
are not calculated. Likewise, electrostatic 1-2 or 1-3 interactions
between the electroneutral oscillators and core charges are also not
calculated. However, electrostatic 1-3 interactions between two
electroneutral oscillators associated with OS1 and OS2 are calculated
according to a Thole screening function. More information in Harder,
E.; Anisimov, V. M.; Whitfield, T.; MacKerell, A. D. & Roux, B.
Understanding the Dielectric Properties of Liquid Amides from a
Polarizable Force Field J. Phys. Chem. B, 2008, 112, 3509-3521



On 29/08/13 11:07, Ana Celia Vila Verde wrote:
> Hi everyone,
> I'm using NAMD to run simulations where polarizability is explicitly
> included through the Drude model. According to
> " The screened Coulomb correction of Thole is calculated between
> atom-Drude pairs that are otherwise excluded from nonbonded interactions."
> To verify that intramolecular electrostatic interactions were indeed
> always calculated between 1-3 Drude pairs when "exclude1-3" was
> selected , I ran two test cases using a simple model of the sulfate
> ion (see and I found some
> puzzling results. The oxygens in the model are polarizable, i.e. they
> consist of a pair of Drude charges, but the sulfur is not.
> Case 1 (attached file a single sulfate ion in its
> equilibrium configuration (i.e., when S is at (0,0,0), the net dipole
> moment of the ion is 0)
> In this case, the electrostatic energy of the system is zero,
> indicating that intramolecular electrostatic interactions are NOT
> calculated, contrary to what is stated in
> Case 2 (attached file a single sulfate ion identical
> in most aspects to that used in Case 1. The only difference now is
> that the drude particles are not in their equilibrium position, so the
> net dipole moment of the sulfate is not zero.
> In this case, the electrostatic energy of the system is NOT zero.
> However, it is also not what it should be if NAMD were calculating the
> interactions between intramolecular drude pairs using Thole screening
> functions. So, NAMD is calculating some sort of intramolecular energy
> in this case, but what?
> Can anyone explain exactly how NAMD calculates intramolecular
> electrostatic interactions between Drude pairs? The files I used to
> run the two test cases are attached in case anyone wants to look into
> this more closely.
> Best,
> Ana

This archive was generated by hypermail 2.1.6 : Tue Dec 27 2016 - 23:21:34 CST