Re: FEP substrate transformation with common moiety

From: Chris Harrison (
Date: Fri Aug 27 2010 - 21:47:26 CDT

Dear Sebastion

Option A will work. That is how dual topology FEP is done ... which
BTW is the assumed (though not mandatory) approach for FEP in NAMD.

However, the resulting path A <--> B is technically a nonphysical
path. If you are interested in the path for such a mutation, one must
ask the question: why are you interested in a path that doesn't exist
in reality, no matter how you construct it..

Option B is problematic. Despite restraints, the rest of the system
will still interact with and respond to the presence of the restrained
portion, A or B. Again, a nonphysical path. Further, you are not
allowing the restrained molecule to relax its geometry WRT its
environment X, and likewise X's geometry won't have the opportunity to
properly relax in response to A/B... so the resulting dynamics will
not be representative of the true potential energy surface, but
instead just wrong dynamics that to the naked eye might look fine.
Few things are worse than a wrong answer that no one knows is wrong.

I'm guessing you are interested in the difference in structure or
dynamics of X bound to A vs bound to B. If so, you're best option
will be to use FEP and simply sample well at both ends of the
perturbation, at state A and state B. You'll get the dynamics, as
well as the free energy change.


Chris Harrison, Ph.D.
Theoretical and Computational Biophysics Group
NIH Resource for Macromolecular Modeling and Bioinformatics
Beckman Institute for Advanced Science and Technology
University of Illinois, 405 N. Mathews Ave., Urbana, IL 61801                            Voice: 217-244-1733               Fax: 217-244-6078
On Fri, Aug 27, 2010 at 8:35 PM, Sebastian Stolzenberg
<> wrote:
> Dear All,
> I would like to use FEP to slowly transform one substrate A, bound to a
> protein in binding site X, into another substrate B bound to X, i.e. not
> considering the exact free energy value of this transformation.
> Initial and final substrates have a common moiety.
> I see two options:
> a) Define a FEP hybrid topology entry A2B with the joint common moiety
> b) Simply include A and B as two separate PDB entries, restrain B to A and
> to X
> Could anybody kindly share their experience what option might work better?
> Thank you very much,
> Best,
> Sebastian

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:56:05 CST