Dear Marawan,
You need to address several problems with QM/MM simulations. These are not directly related to the protocol used for the charge derivation of classical force fields. You need to choose:
a) a sufficiently large QM region.
b) an appropriate QM method (Hamiltonian and basis set) that is able to accurately describe the system/process you are interested in. Density functional theory will be a good choice for many cases.
c) an embedding scheme that properly describes QM-MM interactions. AMBER implements two simple approaches, i) mechanical embedding in which the electrostatic interactions are entirely treated at the classical level using MM point charges; in this case the QM charge density will not get polarized by the environment, ii) electrostatic embedding in which the electric field due to MM point charges is included in the QM Hamiltonian; this will lead to artifacts at the QM/MM boundary which is one of the reasons to choose a QM region as large as possible.
If you want a good representation of the solvation structure within a QM model, you will most likely have to include part of the solvent surrounding the solute in the QM region. There is no straight forward way to do this but adaptive QM/MM methods that make such simulations possible are under development. For details of effects on reaction profiles and radial distribution functions have a look at this paper: http://dx.doi.org/10.1021/ct300331f
All the best,
Andy
On Sep 25, 2012, at 11:48 PM, marawan hussain wrote:
> Hi,
> I want to use the AMBER-Gaussian QM/MM approach to simulate the first solvation shell around a nucleoside..I want to start from an MD structure but i'm worried about the charge conflict issue..For different forcefield AMBER uses different scheme although at the end all may be near because of antechamber, but if i start MD with a definite charge value, from for example DFT like the ff03 FF, if i perform QM/MM the charge of the QM region will be much different as at this case antechamber is not there to balance the charge...or, do i need to choose a certain charge scheme such as MK to derive and to use in the context of the QM-MM interaction...Also, can i choose a DFT functional for the QM region such as B3lyp although my simulation has been already prepared with HF according to the ff99SB scheme.....
>
> Marawan
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
--
Dr. Andreas W. Goetz
Assistant Project Scientist
San Diego Supercomputer Center
Tel : +1-858-822-4771
Email: agoetz.sdsc.edu
Web : www.awgoetz.de
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Sep 26 2012 - 11:30:04 PDT
