Re: Alchemical Free Energy Perturbation

From: Jérôme Hénin (
Date: Thu Mar 02 2006 - 07:39:16 CST

Hi Jonathan,

> I'm trying FEP for the first time with NAMD. I'm
> trying to replicate the tutorial with the ala-ser-ala --> ala3 mutation
> (although in my case I've blocked the ends slightly differently).
> I built the dual .pdb file by hand, generated the .psf file using CHARMM,
> then used alchemify to generate the new .psf file (the one I've attached).
> My questions:
> 1.Does the coordinate file used in the NAMD script have the dual atoms? If
> so, how does it differ from the .fep file?
Yes, it does, so it may be the same as the FEPfile. As a matter of fact, the
default value of the FEPfile parameter is the NAMD coordinate file. The
FEPfile need not have the .fep extension - alhough the extension may be
useful to remember that this particular PDB file contains information about
mutating atoms.

> 2.Am I correct that the .fep file (with the -1,0, and 1 flags) contains all
> atoms in the system? Or should it only be the resiude that is being
> mutated,as you show in the tutorial?
This could be a bit more clearly stated in the manual, and the tutorial should
mention that only a section of the fepFile is shown. FEPfile does contain all
atoms in the system, and all atoms not affected by the transformation should
have a value of 0 in the fep column.

> I've attached the files I'm using. Its running, but I want to check if you
> think I've gotten the protocol correct.
Here are a few remarks about your config file:
1) This system (a single molecule in vacuum) should rather not be simulated
with periodic boundary conditions, and in any case, you don't want to use a

2) The minimization steps are a bit unusual. The simulation should be stable
without these, and they increase the equilibration time because you have to
thermalize the system back to 298 K at each stage. Additionally, 1000 steps
is a very thorough minimization for such a small system. I think you rather
need a short minimization/equilibration run before the actual FEP simulation.

4) In your FEPfile, the C and O atoms of the STA residue are wrongly tagged as
appearing. When I setup FEP files, I usually visualize them in VMD using
"Beta" as a coloring method, so that mistakes in the definition of appearing
and vanishing groups become obvious.

Now I have to say we suspect that something was wrong in the FEP simulations
that were done for the tutorial a long time ago. I will issue a revised
version as soon as I find the time to do so (I know I have been promising
this for months now...)


Jérôme Hénin
Equipe de Dynamique des Assemblages Membranaires
Université Henri Poincaré / CNRS
Tel : (33) 3 83 68 43 91        Fax : (33) 3 83 68 43 87

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:41:41 CST