**From:** Alex Balaeff (*abalaeff_at_polarisqb.com*)

**Date:** Tue Jul 07 2020 - 18:44:21 CDT

**Next message:**Maximilian Ebert: "Dynamic switching GPU/CPU minimization"**Previous message:**Francesco Pietra: "Re: Re: scripts to generate initial boxes of solvents"**Messages sorted by:**[ date ] [ thread ] [ subject ] [ author ] [ attachment ]

Dear NAMD community:

I'd appreciate any advice on conducting pair interaction calculations

in NAMD on a QM/MM system. My calculations reveal an apparent bug --

or am I doing something wrong?

In essence, I am attempting to calculate the interaction energy

(electrostatic and van der Waals) between a ligand (described by QM)

and a solvated protein (described by MM).

For the whole system, the energy looks quite reasonable:

E_elec = -932364.2484, E_vdw = -427.4665

whereas for pair interaction between the QM and MM parts, the vdW

component goes haywire:

E_elec = -118280.8276, E_vdw = 48861334.1618.

If I compute the self-interaction of the MM part I get:

E_elec = -932294.1036, E_vdw = -427.4665

which is reasonable again, except E_vdw is *exactly* the same as for

the complete system. Which is wrong, isn't it? because NAMD does

calculate the vdW interactions between the QM and the MM parts, so

that the total E_vdw should include E_vdw(QM-MM) in addition to

E_vdw(MM-MM). And yes, I do have the vdW parameters for the QM atoms

set up by CGenFF, and don't scale them down to zero.

The self-interaction of the QM part is crazy again:

E_elec = -118339.2930, E_vdw = 48861334.1618

which makes even less sense: the vdW interaction between the QM atoms

is simply not computed, so shouldn't it just be zero??

When I remove the ligand from my MD snapshot and calculate the energy

of the remaining part (without the pair interaction this time), I do

get reasonable values again:

E_elec = -932294.1034, E_vdw = -382.2416

However, why are these values different from those above computed for

the MM self-interactions? Shouldn't those be same values?

Any advice on this confusing situation would be greatly appreciated. I

am attaching a README file describing the problem in more detail, and

put a complete test case (including the NAMD scripts and their log

file outputs) on my Google Drive:

https://drive.google.com/drive/u/1/folders/1TqcdH_rY6J_Vpmfk8sHpFNmoq4TW2ars

Thank you very much,

Alexander.

-- ----- Dr. Alexander Balaeff Polaris Quantum Biotech www.PolarisQB.com (919)-270-5772

- text/plain attachment: README.txt

**Next message:**Maximilian Ebert: "Dynamic switching GPU/CPU minimization"**Previous message:**Francesco Pietra: "Re: Re: scripts to generate initial boxes of solvents"**Messages sorted by:**[ date ] [ thread ] [ subject ] [ author ] [ attachment ]

*
This archive was generated by hypermail 2.1.6
: Thu Dec 31 2020 - 23:17:13 CST
*