From: Jérôme Hénin (jhenin_at_ifr88.cnrs-mrs.fr)
Date: Wed Feb 09 2011 - 03:25:59 CST
Hi,
You can make the assumption that the force distribution and
autocorrelation time are similar across different windows (this is
certainly true for the distribution - it is more problematic for the
correlation time). Then, you just need one additional ABF simulation
where you collect the system force trajectory ; from there you can
estimate the distribution and autocorrelation time.
Jerome
On 9 February 2011 03:58, stefhoor <fabracht1_at_gmail.com> wrote:
> I have performed an ABF calculation by divinding my reaction path into
> numerous intermediate steps. Each interval (0.4 angstrom wide) with a
> pull step size of 0.1 A was calculated in a 2 ns window. At the end I
> gathered everything with a 0 md step configuration file, which gave me
> an integrated pmf file with my entire path. I needed to make an error
> analysis of my calculation. I read several papers, most of them were
> referenced here in namd's mailing list, and got to the conclusion that
> I need to calculate the instantaneous force by adding the
> outputSystemForce option within a colvar {} block. Does this mean that
> I would have to perform all simulations again in order to compute this
> force or is there a way to get them from my current files?
> Thank you
>
>
This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:19:47 CST