Re: FEP simulations?

From: Chris Chipot (
Date: Mon Jan 05 2009 - 05:24:00 CST

Dear Christopher,

> 1) In the paper below, how many steps were performed (eg How much
> sampling?) and what is an ideal mix of accuracy and efficiency? I am
> using the following FEP parameters to perform a double annihilation of
> two simple proteins dimerizing:

One hates by and large to give numbers of steps that guarantee
appropriate convergence. I could say 1,000,000 per λ-state, but
it would not be very satisfactory, would it? These numbers are
in essence system-dependent.

Instead, I do recommend that you probe convergence of your run
by monitoring the different P(ΔU), in addition to how fast your
ΔG reaches a plateau for the various λ-states. This is essentially
spelled out in the text mentioned by Jerome.

> 2) This question will tell you that I am still a novice, but why are
> rigid bonds required to run the FEP?

There is no reason why chemical bonds ought to be frozen, except
perhaps to eliminate high-frequency, hard degrees of freedom and
switch to larger time steps. A small time step might be required
upon igniting the intermolecular interaction of the alternate
topology with its environment, but this is again system-dependent.



Chris Chipot, Ph.D.
Equipe de dynamique des assemblages membranaires
Unité mixte de recherche CNRS/UHP No 7565
Université Henri Poincaré - Nancy 1 Phone: (33) 3-83-68-40-97
B.P. 239 Fax: (33) 3-83-68-43-87
54506 Vandoeuvre-lès-Nancy Cedex

              The wretch, concentred all in self,
              Living, shall forfeit fair renown,
              And, doubly dying, shall go down
              To the vile dust, from whence he sprung,
              Unwept, unhonored, and unsung. Sir Walter Scott

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:52:12 CST