AMBER Archive (2007)

Subject: Re: AMBER: help in QM/MM simulation

From: Gustavo Seabra (gustavo.seabra_at_gmail.com)
Date: Wed Dec 19 2007 - 08:42:19 CST


Hi Pankaj,

What you want is probably an "extension" of (1). Basically, the
procedure we use is:

(1) Minimize / heat / equilibrate system with pure classical
mechanics. I recommend ff99SB or ff03.
(2) Turn on the QM, and run an extra of about 100ps equilibration, to
relax the system to the new Hamiltonian
(3) Production run with QM/MM

Notice that, depending on your system, step (2) may take longer. In
some cases, it may be necessary to initially apply restraints to the
QM part using the final MM snapshot as reference, then slowly remove
the restraints (multiple runs of ~100ps with decreasing restraint,
until you finally eliminate the restraint altogether.) The specific
details of the procedure will depend on what works for your system.

HTH,
Gustavo.

On Dec 13, 2007 4:02 PM, Pankaj R. Daga <pdaga_at_olemiss.edu> wrote:
> Dear Amber Community
>
> I have a very stupid question. I want to perform QM/MM simulation on one
> protein-ligand system. The system, I have been dealing with, needs water
> molecules for the reaction chemistry in the active site. I have been
> considering three different approaches for the same.
>
> 1. Minimize and equilibrate the system with MM and then directly start
> with QM/MM production run
>
> 2. Minimize with MM and further processing with QM/MM potentials.
>
> 3. Start QM/MM processing directly from beginning.
>
> Although, Approach-3 looks to be very convenient, at early stage it is very
> difficult to pin-point the water molecules, to be included in QM region.
>
> I am very new to QM/MM simulation. I would be grateful to you if you could
> suggest me the right approach.
>
> Regards
>
> Pankaj
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu