AMBER Archive (2006)

Subject: Re: AMBER: Gibbs PMF Calculation

From: Carlos Simmerling (
Date: Fri Feb 03 2006 - 05:52:56 CST

no, it means that the windws need to be close enough that
the fluctuations in each window give some of the same values
at the fluctuations in the neighboring window. if you histogram
the value that you are restraining, the histograms should overlap.
they isn't a simple kind of calculation, make sure that you
have carefully read paper on this before you try it.

Atsutoshi Okabe wrote:
> Dear Carlos Simmerling
> Thank you for advising me.
> Then should I use sander module on amber8 manual,i.e. "Potenials of
> mean force using umbrella sampling" and WHAM method ?
> On this manual, it is said “These different simulations(often called
> “windows”) must have some overlap”
> Does it mean that after one simulation is carried out, the next one
> should be started at the structure caluculated by last simulation ?
> A. Okabe
The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to