AMBER Archive (2002)

Subject: Re: dielectric constant for partialy solvated MD

From: Carlos Simmerling (carlos.simmerling_at_sunysb.edu)
Date: Wed Jul 31 2002 - 15:54:12 CDT

if you are interested in the behavior of the solvated portion, and the
rest of the protein is restrained (or else it will likely be unstable),
then a dielectric of 1 makes the most sense to me. This is because
the moving part has explicit solvation.
carlos

----- Original Message -----
From: "Miyashita, Osamu" <omiyashita_at_ucsd.edu>
To: <amber_at_heimdal.compchem.ucsf.edu>
Sent: Wednesday, July 31, 2002 2:55 PM
Subject: dielectric constant for partialy solvated MD

> Hi,
>
> I would like to know your opinions about the dielectric constant for MD.
>
> For MD in vacuum, I think people often use distance dependent dielectric
> constant, such as e=2r or e=4r. If MD is fully solvated, we use e=1, I
> think this is common sense (Maybe not).
>
> In my simulation, I would like to partially solvate only a part of large
> protein, and put CAP to fix water molecules. For example, for a protein
of
> 100x100x100 Angstrom size, I solvate around catalitic site by water sphere
> of radius=25 Angstrom.
>
> Which dielectric constant is appropriate for this case, do you think? Of
> course it will depend on parameter set. I'm interested in AMBER94
parameter
> set.
>
> Thanks,
>
> Osamu Miyashita
> UCSD Physics
>
>