AMBER Archive (2000)

Subject: Re: Monte-Carlo

From: Jed Pitera (jed_at_igc.phys.chem.ethz.ch)
Date: Thu Dec 07 2000 - 11:50:02 CST


Andrew,

        The CMC/MD functionality of AMBER only allows MC moves for
"chemical sampling" -- i.e. sampling different potential functions -- and
not for coordinate sampling. If you want something like a dihedral grid
search then it's possible to use the "interface" front end to sander to do
torsion driving. You could always use some other program for the
coordinate search and then minimize the resulting structures with AMBER;
however, since it sounds like you are keeping a lot of the system rigid
this might not be the most efficient approach.

        Simulated annealing with locally enhanced sampling (LES) is
probably the AMBER functionality best suited to the problem you describe,
assuming you're trying to find the global minimum energy conformation of
your subsystem.

--Jed

Jed Pitera, PhD Postdoctoral Researcher
van Gunsteren Group, Physical Chemistry
ETH Zurich, Switzerland

On Wed, 6 Dec 2000, Andrew Aird wrote:

> Hello all
>
> I'd like to do a conformational search on a fairly large system consisting of
> three protein-substructures. Only part of each substructure is allowed to move.
> I'm not sure, if CMC/MD is the right choice for this kind of job. Has anyone
> worked on a similar problem??
> Any comments would be of help...
>
> Tahnks in advance
>
> Andrew
> --
>
>
>
>
>
> ---------------------------------------------------
> Andrew Aird
> 3.Physikalisches Institut
> Universität Stuttgart
> Pfaffenwaldring 57
> D 70569 Stuttgart
> e-mail: a.aird_at_physik.uni-stuttgart.de
> ---------------------------------------------------
>