Re: Replica exchange, colvars & harmonic restraints

From: Giacomo Fiorin (
Date: Mon Jul 29 2013 - 16:22:56 CDT

Hello Andrei, your configuration looks correct to keep the distance between
the protein's COM and the origin (or the center of your choice).

Regarding the two values, obviously you want the center of the harmonic
potential to be zero, to keep the distance low.

Regarding the force constant, it depends on how big is your protein or
peptide: start with 10, and if peptide still drifts too much for your
taste, increase it until the drift stops.


On Mon, Jul 29, 2013 at 2:36 PM, Andrei Tudor <>wrote:

> Hello,
> I want to run a replica exchange simulation with my protein system. I am
> pretty sure I got the configuration of the replica exchange part working,
> but I want to use colvars and harmonic restraints to keep the protein in
> the center of the box (using it's center of mass).
> I have never used colvars and I find them quite confusing. I know I have
> to create a group for the atoms forming my protein and use somehow the
> center of mass of that selection.
> I want to use harmonic restraints to let the center of mass oscillate a
> bit, but not by much. Let's say 5 A. How does one make the config file of
> the colvars?
> I am thinking of doing it something like this:
> colvar {
> name COM
> distance{
> group1 {
> atomNumbers
> }
> group2 {
> dummyAtom (center of initial coords of the peptide)
> }
> }
> }
> harmonic {
> colvars COM
> centers 0 ?????
> forceConstant 10.0 ???????
> }
> Does this look any good? Where I put the question marks, it means that I
> don't know why to put the values I put. I don't really understand the
> harmonic options.
> Also if I calculate the coords of the center of the peptide in vmd, is
> that what I should put as the dummyAtom coordinates?
> Any help will greatly appreciated.
> Thank you,
> Andrei

This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:21:29 CST