Re: [AMBER] sander QM/MM MD

From: Dr. Andreas W. Goetz <>
Date: Wed, 20 Jul 2016 11:23:30 -0700

Dear Masoud,

> I am using AMBER 12 to do QMMM calculation using G09 as external QM
> package. I have a couple questions and I would be grateful
> if somebody could help.
> - 1) Is it electronic embedding? The user manual (Amber 12) state that "*For
> QM/MM simulations with electronic embedding it is mandatory to specify the
> charge of the** QM region in the &qmmm namelist with variable qmcharge
> in addition to specifying the charge** in the namelist that is**
> specific for the external QM program". *But I received the message that
> these keywords are deprecated (charge in the case G09)*.* The Gaussian
> input "*#P M062X/6-31g** SCF=(Conver=8) NoSymm Force Guess=Read Charge
> Prop=(Field,Read)*" contains the *charge* keyword and the coordinate/MM
> charges of the surrounding. I think this means the electrostatic part of
> environment is included in the QM Hamiltonian and the calculation is EE. I
> would be grateful if somebody could confirm this.

It seems that you are not using Amber 12. Charge and spin multiplicity for external QM programs are defined in the qmmm namelist since version 14 (and their definition in the namelist for the external QM program thus deprecated). Please make sure that you use the manual that corresponds to the version of Amber that you are using.

Since version 14 the sander program is part of AmberTools, which is free. You probably should upgrade to the latest release version 16.

Your QM/MM calculation is using electronic embedding , including the electrostatic potential of the point charges in the environment in the QM Hamiltonian.

> - 2) How *EXTERNESCF* is realated to SCF in G09? I am also confused
> about the sander output. For a point in my calculation I get,*
> EXTERNESCF= -216070.1690* while the G09 output for same point is *SCF
> Done: E(RM062X) = -1704.92860364*. The G09 energies are A.U which if
> converted to kcal/mol or kj/mol would be very different from the value
> given as *EXTERNESCF. *what is this discrepancy?

Gaussian includes the electrostatic energy among the point charges in its output. ESCF is the SCF energy without this contribution, which is reported as "Self energy of the charges” in the Gaussian output.

All the best,

Dr. Andreas W. Goetz
Assistant Project Scientist
San Diego Supercomputer Center
Tel: +1-858-822-4771

AMBER mailing list
Received on Wed Jul 20 2016 - 11:30:02 PDT
Custom Search