Re: Request for information regarding ABF in NAMD.

From: Jérôme Hénin (
Date: Fri Dec 11 2009 - 08:51:35 CST

Dear ArunKumar,

I am copying this message to the NAMD mailing list, as it could be of
interest to other readers.

I can see several possible ways of approaching your problem. One is to
define two RMSD variables (wrt each of the end points), and study them
in 2d. A related method would use one variable defined as the
difference between two RMSD values (colvar components). The latter
method is illustrated in our latest publication:
Finally, you could compute a displacement vector between the two
conformations, and then follow the transition by acting on the
position of the system along this displacement vector (that is handled
by the colvar component called "eigenvector").


2009/12/11 Arun Kumar Subramanian <>:
> Hi,
> I am a postdoc in Prof. Gunther Peters lab in the Technical University of Denmark, Copenhagen. I am trying to employ ABF calculations to understand the molecular movements leading to opening and closing of a protein loop (around 10 residues in length). One thing I couldnt figure out is that is it possible to tell the initial and final positions of the loop (because we have crystal structures of the protein open and closed loops) and direct the force biasing in this direction. I could see that I can set different types of restraints like distance, torsions, etc. such that the force averaging is done on the defined set. But, I couldnt understand if we can tell the direction in which the forces should act on. We are unable to produce the intermediate states of loop more accurately if we want to conduct the simulation in several bins and try to see if there is any energy barrier. Hence it would be great if there is a way we could mention the closed and open states and figure out what movements are crucial to e
nable them.
> Thanking you.
> Sincerely,
> ArunKumar. S.

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:53:35 CST