Re: [AMBER] Three body potential

From: Jason Swails <>
Date: Fri, 13 Sep 2013 15:15:39 -0400

On Fri, Sep 13, 2013 at 2:47 PM, MOHD HOMAIDUR RAHMAN

> Dear Amber User
> I want to use three body potential in amber force field. Please guide me to
> how I will proceed this.

There's already a three-body potential in Amber -- the angle term. (And a
4-body term for torsions and improper dihedrals).

If you want a 3-body nonbonded term, you will need to modify the source
code. My suggestion is to see how the current pair potentials are handled
(this is more complicated in PME, since you have periodic boundary
conditions and an imaged pairlist) and try to extend that code to your

The short_ene.F90 file has many routines for the existing energy terms --
see how the terms are handled there and copy what you can reuse.

Another issue is how to get your parameters into sander. You can either
modify the topology file somehow (e.g., ParmEd), hard-code your parameters
into the sander source code, or put them in a separate file that is read
in. There are advantages and disadvantages to each approach.

Of course, you can also use another program that may make it easier to add
your own force routines -- like maybe OpenMM or NAMD through the tcl
interface). For Amber, you will have to familiarize yourself with the
source code and likely answer most of your questions via experimentation
and tests. When you go down this route, it is much harder to provide
assistance, so you'll likely be debugging stuff on your own. If you do
have a question, it helps to be as specific as possible about what you
tried, what happened, what you expected to happen, and maybe what you want
to happen.

Good luck,

Jason M. Swails
Rutgers University
Postdoctoral Researcher
AMBER mailing list
Received on Fri Sep 13 2013 - 12:30:02 PDT
Custom Search