Re: [AMBER] Positional restraints

From: Jason Swails <>
Date: Wed, 13 Nov 2013 16:18:59 -0500

On Wed, Nov 13, 2013 at 3:41 PM, George Tzotzos <> wrote:

> Hi everybody,
> I'm simulating a system in which there is a conserved water molecule in
> the binding site of the protein. I used restraint_wt=1.0 during energy
> minimization, restraint_wt=0.1 during heating and density equilibration.
> This resulted to the water molecule flying out of the binding site.

I wonder if this is an imaging artifact. These are actually fairly strong
restraint weights for dynamics (assuming that r1 and r4 are significantly
smaller and larger than r2 and r3, respectively). Even with a restraint
weight of 0.1, there is about a 1 kcal/mol penalty for a single particle to
move 3 away from the reference position.

Are you applying the positional restraints to the whole system or just to
the conserved water molecule?

It's obvious that I should had used stronger restraints. Is there a rule of
> thumb as to the range of restraints that one should use during
> minimization, heating and density equilibration? Should there be a gradual
> relaxation of restraints before production MD?

If you ultimately plan on getting rid of the restraints, then what you used
is typical compared to what I use.


Jason M. Swails
Rutgers University
Postdoctoral Researcher
AMBER mailing list
Received on Wed Nov 13 2013 - 13:30:03 PST
Custom Search