RE: ABF atom selection

From: Parisa Akhski (
Date: Tue May 10 2011 - 09:16:29 CDT

Hi Jerome,

Thanks for your reply. I think it is better to use my own system and explain what I did so that I could realize what the problem is.

I am running ABF simulation for Ammonium ion in the channel.

I did the simulation for the same system in two different ways:

1. I define ammonium ion in the colvars as below:

atomnumbers { 1 2 3 4 5}

in which number 1 is the nitrogen atom and the rest are hydrogen atoms.

2. I define the ammonium ion in the colvars using only nitrogen atom:

atomnumbers { 1 }

in which 1 is the nitrogen atom number.

In both cases I confine the ion movement in XY plane using "distanceXY".

>From what you said, I expect to get the same results for these systems. However, they are different and only the results of the second case, make sense for my system.

I am wondering what is wrong with my definition and why the results are diffident?



-----Original Message-----
From: on behalf of Jérôme Hénin
Sent: Sun 5/8/2011 12:41 PM
To: Parisa Akhski
Subject: Re: namd-l: ABF atom selection
Hi Parisa,

> I am obtaining PMF profile using ABF method (NAMD 2.7 colvar). In "hydration
> free energy calculation for methane" in the tutorial, atom selection is
> defined using atom numbers {1 2 3 4 5}. How this atom selection is different
> from defining only carbon atom number since carbon is the COM for this
> molecule? I am assuming these two are the same.

Almost, but not exactly. Using all atoms to define the group will
collect more data, leading to faster convergence.

> However, if I want to apply force to confine methane movement in XY plane
> (using distance XY), this is not true any more. As I understood, in the
> first case, force is applied to all atoms individually (not COM of these
> atoms) whereas in the second case, force is only applied on the carbon atom.

That is not correct. For both coordinates, you can use either
individual atoms or groups. If a force is "applied to a group" or
"applied to the COM of a group", it means that it is distributed over
all atoms in the group.

Best wishes,

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 05:23:54 CST