Re: WHAM analysis for Umbrella sampling

From: Aron Broom (broomsday_at_gmail.com)
Date: Wed May 01 2013 - 14:12:23 CDT

assuming the order of the columns in your file is just: timestep,
variable_of_interest, other stuff... then you don't need to do any
extraction, just feed it the raw files.

On Wed, May 1, 2013 at 2:59 PM, Ada Zhan <zyqfrog10_at_msn.com> wrote:

> Aron,
>
> I have .colvars.traj files and the output frequency is high enough (every
> 5ps). The easiest way for me would be to extract the information from
> .colvars.traj directly and apply them into WHAM by Grossfield.
> Thank you!
>
> Chris,
>
> Thanks for your encouragement! I would like to try to write up such kind
> of code for learning purpose. For now, a well established software is more
> straight forward and safe for me. :-)
>
> Thanks again!
>
> Best regards,
> Ada
>
> ------------------------------
> Date: Wed, 1 May 2013 14:09:53 -0400
>
> Subject: Re: namd-l: WHAM analysis for Umbrella sampling
> From: broomsday_at_gmail.com
> To: chipot_at_ks.uiuc.edu
> CC: zyqfrog10_at_msn.com; jerome.henin_at_ibpc.fr; namd-l_at_ks.uiuc.edu
>
>
> Ada,
>
> Do you not also have the .colvars.traj files?
>
> I take Jerome's point about the discreet nature of the data, but if the
> frequency is far higher than the decorrelation time than it's mostly
> irrelevant.
>
>
> On Wed, May 1, 2013 at 1:58 PM, Chris Chipot <chipot_at_ks.uiuc.edu> wrote:
>
> Ada,
>
> WHAM is not your only option to reconstruct your potential of mean force.
> If you are dealing with a one-dimensional PMF, you can write up a quick
> and dirty code that patches together the free-energy profiles obtained in
> each stratum from the biased probability distributions. This code would
> minimize the difference in curvature between adjacent profiles in their
> overlapping region. If, on the contrary, the dimensionality is higher, you
> are pretty much forced to go for multidimensional WHAM.
>
> Good luck,
>
> Chris Chipot
>
>
>
> On 5/1/13 7:20 PM, Ada Zhan wrote:
>
> Chris,
>
> Thanks for your suggestions! I will take a look at those papers.
> I understand the part that WHAM equations require only histograms or
> counts or probability distribution.
> I just could not find a standalone and well established WHAM software that
> accepts histograms directly. For WHAM by Grossfield, it was confirmed by
> the developer that I do need to modify the code a little bit to do that.
> Modifying the code or even writing a code like this is too much for me.
> If someone can develop a WHAM plugin for VMD that bridges the histogram
> files from NAMD and WHAM analysis, it would a great plus!
>
> Thank all of you again! You guys are really helpful!
>
> Best regards,
> Ada
>
> ------------------------------
> Date: Wed, 1 May 2013 09:01:44 +0200
> From: chipot_at_ks.uiuc.edu
> To: zyqfrog10_at_msn.com
> CC: jerome.henin_at_ibpc.fr; broomsday_at_gmail.com; namd-l_at_ks.uiuc.edu
> Subject: Re: namd-l: WHAM analysis for Umbrella sampling
>
> Ada,
>
> I suggest that you have a look at the original Kumar et al. paper from 1992
> (and possibly the earlier work of Bob Swendsen). There, you will see that
> the only thing needed to solve the WHAM (Ferrenberg-Swendsen) equations
> is your biased probability distributions in the different strata of your
> reaction
> path and the biasing potentials that you have used in these different
> strata.
> Period. This is also spelled out in Benoit Roux's 1995 paper on potential
> of
> mean force calculations.
>
> Chris Chipot
>
>
>
> On 5/1/13 12:18 AM, Ada Zhan wrote:
>
> Hi there,
>
> Thanks for all your responses!
>
> Jerome, are you using WHAM from Grossfield lab? That is what I am using.
> According to their WHAM manual, time series files are required to contain
> the information of time and positions. And their commands process the time
> series files as the starting point.
> I understand they are processing the time series files into histograms or
> counts that will be implemented into the WHAM equations.
> My curiosity would be how to direcetly implement NAMD generated histograms
> into existing WHAM software without modifing their codes.
>
> Thanks!
>
> Best regards,
> Ada
>
> > Date: Tue, 30 Apr 2013 11:07:52 +0200
> > From: jerome.henin_at_ibpc.fr
> > To: broomsday_at_gmail.com
> > CC: namd-l_at_ks.uiuc.edu; zyqfrog10_at_msn.com
> > Subject: Re: namd-l: WHAM analysis for Umbrella sampling
> >
> > Hi,
> >
> > I recommended the histogram solution to avoid the downsampling that
> happens in the trajectory. Apparently your implementation of WHAM requires
> a trajectory, I am not sure why. One possible advantage of saving a
> trajectory is that it is not discretized in colvar space, so you can
> optimize the binning width afterwards. If the histogram is fine enough
> though, that's not a problem.
> >
> > Cheers,
> > Jerome
> >
> >
> > ----- Original Message -----
> > >
> > > yeah you don't do the histogram thing. You just want the
> > > .colvars.traj files.
> > >
> > >
> > >
> > >
> > > On Mon, Apr 29, 2013 at 3:41 PM, Ada Zhan < zyqfrog10_at_msn.com >
> > > wrote:
> > >
> > >
> > >
> > >
> > > Hi there,
> > >
> > > I want to perform WHAM analysis on histogram files generated from
> > > umbrella sampling simulations.
> > >
> > > According to WHAM instructions, time series files contain the
> > > information of time and position that are recorded in the .traj
> > > files.
> > > Now I used 'histogram' function provided by colvars. How do I use
> > > these histogram .dat files for WHAM use? Do they serve as time
> > > series files also? Or we use them somewhere differently? If we put
> > > them in the field of time series files, how would WHAM process them
> > > since the meaning of the provided information is totally changed?
> > >
> > > Thanks!
> > >
> > > Best regards,
> > > Ada
> > >
> > >
> > >
> > > --
> > > Aron Broom M.Sc
> > > PhD Student
> > > Department of Chemistry
> > > University of Waterloo
> > >
>
>
>
> --
> _______________________________________________________________________
>
> Chris Chipot, Ph.D.
> Theoretical and Computational Biophysics Group
> Beckman Institute
> University of Illinois at Urbana-Champaign
> 405 North Mathews Phone: (217) 244-5711
> Urbana, Illinois 61801 Fax: (217) 244-6078
>
> E-mail: chipot_at_ks.uiuc.edu
> Christophe.Chipot_at_edam.uhp-nancy.fr
> Web: http://www.ks.uiuc.edu/~chipot
> http://www.edam.uhp-nancy.fr
>
> The light shines in the darkness, and the darkness has not overcome it.
> John 1:5.
> _______________________________________________________________________
>
>
>
> --
> _______________________________________________________________________
>
> Chris Chipot, Ph.D.
> Theoretical and Computational Biophysics Group
> Beckman Institute
> University of Illinois at Urbana-Champaign
> 405 North Mathews Phone: (217) 244-5711
> Urbana, Illinois 61801 Fax: (217) 244-6078
>
> E-mail: chipot_at_ks.uiuc.edu
> Christophe.Chipot_at_edam.uhp-nancy.fr
> Web: http://www.ks.uiuc.edu/~chipot
> http://www.edam.uhp-nancy.fr
>
> The light shines in the darkness, and the darkness has not overcome it.
> John 1:5.
> _______________________________________________________________________
>
>
>
>
> --
> Aron Broom M.Sc
> PhD Student
> Department of Chemistry
> University of Waterloo
>

-- 
Aron Broom M.Sc
PhD Student
Department of Chemistry
University of Waterloo

This archive was generated by hypermail 2.1.6 : Wed Dec 31 2014 - 23:21:10 CST