Re: [AMBER] reference group and pulled group in umbrella sampling

From: Jio M <jiomm.yahoo.com>
Date: Sun, 2 Jun 2013 01:42:54 -0700 (PDT)

Dear Jason,

> There's no such thing as a 'reference' group.  The umbrella potential
> exerts a symmetric force on both groups that pulls them towards each other
> (or pushes them away from each other) with equal strength.  It's just pure
> Newtonian dynamics.


Thanks for your reply. I am confused how AMBER and GROMACS are working! I have done AMBER umbrella sampling but am confused about GROMACS things as I want to compare one job in Gromacs. There in Gromacs one has to define reference group (with flags pull_group0 and pull_group1). 

Can you please suggest something regarding this or anyone please suggest who has done umbrella sampling in AMBER and Gromacs just for comparison.

regards,
JIomm


________________________________
 From: Jason Swails <jason.swails.gmail.com>
To: Jio M <jiomm.yahoo.com>; AMBER Mailing List <amber.ambermd.org>
Sent: Sunday, June 2, 2013 7:29 AM
Subject: Re: [AMBER] reference group and pulled group in umbrella sampling
 

On Fri, May 31, 2013 at 6:38 AM, Jio M <jiomm.yahoo.com> wrote:

> Dear All,
>
> I want to do umbrella sampling between two molecules. So I have defined
> atoms in igr1 and igr2. Just want to know which one will be considered as
> reference group and which one is pulled?
> Does igr1 becomes reference group by default?
>

There's no such thing as a 'reference' group.  The umbrella potential
exerts a symmetric force on both groups that pulls them towards each other
(or pushes them away from each other) with equal strength.  It's just pure
Newtonian dynamics.

Also if I have one heavy molecule 'A' (number of atoms and also weight)
> compared to other molecule 'B' then should 'A' be considered as a reference
> group? Or it does not matter
>

The heavier 'group' of atoms will move less due to the applied force (since
the acceleration is the force divided by the total mass).  If you think of
a classical spring attached to a mass at one end and an immovable wall at
the other (where the wall is the 'reference'), this is equivalent to that
spring being attached to two objects, with the mass of the 'wall' object
being infinite.

But at the end of the day, it does not matter.  All of the groups move
according to the classical equations of motion where all forces acting on
all particles are conservative (I believe, barring strange cutoff schemes
or inaccuracies of multiple time-stepping).

HTH,
Jason

-- 
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sun Jun 02 2013 - 02:00:02 PDT
Custom Search