Re: ABF question

From: Jérôme Hénin (
Date: Wed Sep 21 2011 - 07:01:50 CDT

Dear Hamid,

Questions like this should be sent to the NAMD mailing list, which I
am adding as a CC. Please subscribe to the list and use it for any
follow-up messages.

About your problem: I understand that you are trying to make ABF push
a ligand into its binding site. This is not easy, as there are usually
entropic barriers to that process. Furthermore, you are using a
"zDistance" coordinate, which is very indiscriminate. It does not
guarantee that the ligand will line up properly for entering the site.
The ligand might be pushing into the protein besides the binding site,
and the z coordinate will not tell you if that happens. This can only
work if the ligand is naturally restrained to a more or less linear
pathway. Also, you have to allow for enough time for sampling changes
in orientation and conformation of the ligand (unless it has only one
atom, as seems to be the case here).

My advice is:
1) ask yourself what exactly you want to measure, and if ABF is the right tool
2) look for the best reaction coordinate to describe the process unambiguously
3) visualize all MD trajectories in VMD and check that everything
happens as you expect


On 21 September 2011 14:17, Hamid Hadi Alijanvand <> wrote:
> Dear Dr. Hénin,
> Greetings.
> I wonder if you answer my question about ABF method implemented in NAMD.
> I setup a PMF for a ligand receptor. I try to enclose the ligand (main
> group) from bulk water to the binding site in the receptor lumen (ref group)
> along Z axis. I use the distanceZ component and e axis (0,0,-1);
> distanceZ {
>       main {
>          atomnumbers { 1 }
>       }
>       ref  {
>          atomnumbers { 2,3,....}
>  }
>        axis  ( 0.0, 0.0, -1.0 )
>    }
> Then I draw the .pmf file. I confused with the results. As the ligand be
> closer to the ligand binding site in the lumen (lower values of RC) the PMF
> value be larger (more positive). Does it mean the ligand has affinity for
> receptor? I think if the axis set to (0,0,1) then the positive pmf value
> means the ligand tends to move towards larger value of RC along Z-axis.
> I would really appreciate you if guide me about these issues.
> Regards.
> Hamid Hadi Alijanvand
>  PhD. student of Biophysics.  Institute of Biochemistry and Biophysics
> (IBB),   University of Tehran, Enghelab Ave. ,  Tehran, Iran  P.O. Box:
> 13145-1384
>  Alt. e mail: h2a1385[at]

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:57:45 CST