Re: Jarzynski equality and the pull direction in SMD

From: Chris Harrison (
Date: Sun Apr 05 2009 - 03:09:54 CDT


To properly capture the free energy describing the conformational space of the binding event, the SMDs would need to include a very large statistical sample of both multiple initial configurations and multiple directions for the "binding vector"; however that assumes the real binding rxn coordinate is exactly linear which is not unreasonable but usually thought to be unlikely. I would strongly suggest that if all you want is the dG or dA of binding, that you pursue instead Alchemical methods like Free Energy Perturbation (FEP) or Thermodynamic Integration (TI) as implemented in NAMD 2.7b1. If however you are interested in the pathway of binding and the resulting PMF describing the free energy for a given binding trajectory, then SMD, Adaptive Biasing Force (ABF), and metadynamics would be methods to use .... but to properly describe the "free energy of the binding trajectory" these three methods will require *enormous* amounts of sampling in comparison to FEP or TI.


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
Paulo Cesar Telles de Souza <> writes:
> Date: Thu, 2 Apr 2009 21:00:40 -0300
> From: Paulo Cesar Telles de Souza <>
> To: NAMD list <>
> Subject: namd-l: Jarzynski equality and the pull direction in SMD
> Return-Path:
> Message-ID: <>
> X-Spam-Status: No, score=-1.8 required=5.0 tests=BAYES_00,HTML_00_10,
> 	HTML_MESSAGE autolearn=no version=3.1.7-0+tcb1
> Dear all,
> I have a doubt about the aplication of SMD simulations to obtain binding
> free energy using the Jarzynski Equality: Does the pull direction have to be
> the same in all the SMD trajectories, but with different initial
> configurations, or  should I vary the pull directions in order to have a
> bigger sample of ways of occurring a process?
> In my study I try to obtain the binding free energy of ligant that is docked
> in the surface of a receptor. I think that the reaction coordinate is almost
> the same if I change a little the pull direction, but I did not find any
> requirement in Jarzynski Equality theory saying that the sample of work has
> to be composed by SMD simulations of only one pulling direction.
> Can anyone help me ?
> Best regards
> Paulo Cesat Telles de Souza
> Campinas State Universiy
> Brazi

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:52:33 CST