To clarify question 3, the van der Waals term (VDWAALS) in QM/MM calculations contains interactions among MM atoms and between QM and MM atoms (that is MM-MM, and QM-MM, but not QM-QM). All from the classical Amber LJ potential.
As Brian pointed out, EELEC contains only electrostatic interactions among charges in the MM region (unless you use DFTB or semiempirical Hamiltonians and use PME). The Coulomb interaction between QM charge density and MM point charges is part of the SCF energy.
All the best,
Andy
—
Dr. Andreas W. Goetz
Assistant Project Scientist
San Diego Supercomputer Center
Tel: +1-858-822-4771
Email: agoetz.sdsc.edu
Web: www.awgoetz.de
> On Apr 25, 2016, at 4:50 PM, Brian Radak <brian.radak.accts.gmail.com> wrote:
>
>
>
> On 04/25/2016 03:46 AM, 王美婷 wrote:
>> Dear all
>> I want to calculate the free energy difference between QM/MM Hamiltonian and MM Hamiltonian for an acetamide molecule soaked in periodic boxes of TIP3P waters. For the QM/MM simulation, the QM region is the solute molecule acetamide and the MM region is water. The QM method used in the QM/MM simulation is BLYP/6-31g* .
> Presumably you mean to get the relative solvation free energies? The
> QM/MM to MM free energy is completely arbitrary and physically
> meaningless. If you are interested (and not already aware) Koenig and
> Brooks have several papers comparing the relative solvation free
> energies of such methods and also have tested several ways of computing
> them from pure MM trajectories.
>
>> I use SANDER module with option imin=5 to calculate the potential energy of each step of the QM/MM MD (with the same parmtop and imaged crd )。 The input file of imin=5 and QM/MM were attached.
>> My quetions are:
>> 1, The potential energy of the same snapshot given by QM/MM MD (which can be read in the output file) is different to that calculated by SANDER with imin=5, the energy of QM region is almost the same, but the other items have large difference especially for the BOND term. I want to know why?
> This should always be the case. The energies in sander are a (half?)step
> behind the trajectory file due to the leap frog integrator.
> Your only choice is to re-evaluate the trajectory within imin=5 twice.
>> 2, I also want to know whether the potential energy printed in the out file of QM/MM MD is the total potential energy of the system or not? Any scale or missing terms used here ?
> See above. There shouldn't be any scaling unless you are looking at QM
> output in hartrees for some reason.
>
>> 3, Is the vdW term printed in QM/MM MD out file is the sum of the vdW interaction energy among the solvent molecules and that between QM and MM region?
> vdW interactions are pure MM terms in AMBER unless you use the (highly
> experimental) QXD module.
>> 4. Is the ELE item printed in QM/MM MD out file only include the electrostatic interaction between QM and MM region?
> I believe (but am not certain) that the QM SCF energy is not included in
> that total. Obviously the SCF energy includes electrostatic components.
> The MM/MM electrostatics are also obviously part of the EELE term.
>>
>> --
>>
>> mt wang
>>
> HTH,
> Brian
>>
>>
>> _______________________________________________
>> AMBER mailing list
>> AMBER.ambermd.org
>> http://lists.ambermd.org/mailman/listinfo/amber
>
> --
> Brian Radak
> Postdoctoral Scholar
> Gordon Center for Integrative Science, W323A
> Department of Biochemistry & Molecular Biology
> University of Chicago
> 929 E. 57th St.
> Chicago, IL 60637-1454
> Tel: 773/834-2812
> email: radak.uchicago.edu
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Apr 26 2016 - 16:30:04 PDT
