LAMMPS WWW Site - LAMMPS Documentation - LAMMPS Mailing List Archives
Re: [lammps-users] Questions about the bias added by metadynamics
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [lammps-users] Questions about the bias added by metadynamics


From: Giacomo Fiorin <giacomo.fiorin@...24...>
Date: Wed, 25 Oct 2017 10:59:25 -0400

Hi Xinyi, that's correct, the biasing forces will be applied only to the atoms that are part of the collective variable's definition.  Nearby water molecules will only experience indirect effects, as the atoms that you are forcing under the metadynamics bias move away.

The "lowerBoundary" and "upperBoundary" options are only needed to define the range of the PMF that you are recording, but do not contribute any additional forces.

If you want to confine the variable within the range 10-30, just add a wall potential, e.g. using an harmonicWalls restraint:
http://colvars.github.io/colvars-refman-lammps/colvars-refman-lammps.html#sec:colvarbias_harmonic_walls

Giacomo

On Mon, Oct 23, 2017 at 11:18 AM, Xinyi Shen <xinyis@...1250...683...> wrote:
Dear Prof. Giacomo Fiorin,

I am trying to use your collective variables module(colvars) for LAMMPS, and I have some questions about it. Could you please help me?

The first question is about the bias added by metadynamics. Are the bias added to the whole system, or only to the target atoms in our system?For example, if I try to use the distanceZ between 2 particles in water as my colvars, when we add bias to the system, will the water molecule also be influenced directly by the added bias? Or the bias are only added to system of the atoms in the two particles?

Another question is about the boundary for metadynamics. I am wondering how the boundary works, I tried to use the boundary to control the system, but I notice it do not work. Still use the example above. I set the boundary of the distanceZ as 10-30, and in fact in the system it can be 10-50, and I notice the system still try to explore 30-50. Does it mean metadynamics cannot control the direction at all (Since we just add bias, it seems hard to decide the direction)? Then why do we need to set the boundary? Or maybe our system can do something like 'Oh! I am out of the boundary now, maybe I need to try something else'?

Thank you very much!

Best wishes,
Xinyi



--
Giacomo Fiorin
Associate Professor of Research, Temple University, Philadelphia, PA
Contractor, National Institutes of Health, Bethesda, MD