Re: Determining Transition State from an Unfolding Simulation

From: E. Prabhu Raman (
Date: Sun Sep 23 2007 - 21:43:36 CDT

Using Snapshots of a trajectory of Kinetics simulation at a constant temperature, a method called Progress Variable Cluster has been used to pin-point the structures of the transition state ensemble(TSE).
REF : Chemical Physics Volume 307, Issues 2-3, 27 December 2004, Pages 251-258
The basic idea being that the passage through the transition state can be identified by the time-point(s) that record a maximal change in a suitable reaction coordinate (example Rg, or number of native contacts)
However,I WOULD CAUTION that this approach,to my best knowledge has been tried for Coarse Grained Model folding studies and more importantly lots(~100) of independent trajectories were used to get a picture of the TSE.
I assume that you might not have too many unfolding trajectories. But since this procedure is easy to apply, you can try it out and compare your TSE from any experimental available data(phi-values).
The reference given above uses a clustering algorithm to cluster similar structures. A first pass at TSE determination could be not to cluster, but simply pick out the structures that record the steepest buildup(or down) of the reaction coordinate and look at the structures to see if it is any meaningful at all.


E.Prabhu Raman
Ph.D Student, Bioinformatics & Computational Biology
George Mason University

----- Original Message -----
From: Neelanjana Sengupta <>
Date: Sunday, September 23, 2007 7:33 pm
Subject: Re: namd-l: Determining Transition State from an Unfolding Simulation

> Hi,
> This would work (if at all) assuming your unfolding pathway
> retraces the
> protein folding pathway. If you figure out a way to determine if
> this is
> what is going on (figuring this out would indeed be non-trivial),
> you may
> then closely examine the timeline of a unfolding parameter (Rg, for
> instance) and see if it is stable over an extended period of time.
> If you
> find that the energy is stable too, then you "may" think that you have
> encountered a transition state. However, a lot of caution is
> required in
> your treatment...
> Would be great if others share their thought too!
> Cheers,
> Neelanjana
> On 9/23/07, Arun Krishnan <> wrote:
> >
> > Hi All,
> >
> > Is there a way to calculate the transition state structure from
> Unfolding> data? From the plot of RMSD vs time maybe?
> > Or is there some other way? Any pointers would be much appreciated.
> >
> > Cheers,
> >
> > Arun
> >
> >
> >
> --

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:45:17 CST