From: Nehad Elsalamouny (nehade_at_uow.edu.au)
Date: Wed Aug 30 2017 - 19:20:09 CDT
On 31 Aug 2017, at 1:26 AM, Brian Radak <bradak_at_anl.gov<mailto:bradak_at_anl.gov>> wrote:
Can you not load an increasing number of files in sequence? For example if you have (sequential) files log1.feplog, log2.feplog, and log3.feplog, can you just not just run parseFEP as:
parseFEP log1.feplog log2.feplog
parseFEP log1.feplog log2.feplog log3.feplog
Again, I might be missing something significant about the way parseFEP works.
Agreement of forward and backward simulations is really only a negative test for convergence. That is, if they don't agree, then they are not converged. If they do agree, then there is no firm statement to be made about convergence. As a way to convince yourself of this, imagine that you ran a simulation with one step per lambda value in the forward and then reverse directions. Since the energy does not likely change much, the two sets of simulations would give nearly identical results, but are pretty clearly not converged.
On 08/29/2017 05:56 PM, Nehad Elsalamouny wrote:
Brian, Thanks for your reply.
That is what I have done. But the dG in that case is not the cumulative. It is the dG for each individual simulation. Here you will find the plot attached (Picture1). Also I plotted dG for forward and backward simulations (Picture2). Can these 2 plots help me know whether my calculation has converged??
Thanks for help,
On 30 Aug 2017, at 1:08 AM, Brian Radak <bradak_at_anl.gov<mailto:bradak_at_anl.gov>> wrote:
I think I understand your question, but I do not have substantial experience with ParseFEP. It might be better if you posed this question to the VMD listserv, since that is a VMD plugin.
As a guess - maybe you can just load the simulation in increments and run separate ParseFEP calculations? I don't know if ParseFEP has subsampling options, although that would be very useful.
On 08/29/2017 12:18 AM, Nehad Elsalamouny wrote:
Dear namd users,
I ran FEP calculation over 20 windows (forward and backward) for 10ns. I used parsefep (BAR) to analyse my data. I want to check the convergence. What I have from my calculation is the dG for each window (i.e. dG (1ns to 2ns, then 2ns to 3ns, and so on) so I need to calculate the cumulative binding free energy. How can I do this?
Thanks in advance
-- Brian Radak Postdoctoral Appointee Leadership Computing Facility Argonne National Laboratory 9700 South Cass Avenue, Bldg. 240 Argonne, IL 60439-4854 (630) 252-8643 brian.radak_at_anl.gov<mailto:brian.radak_at_anl.gov> -- Brian Radak Postdoctoral Appointee Leadership Computing Facility Argonne National Laboratory 9700 South Cass Avenue, Bldg. 240 Argonne, IL 60439-4854 (630) 252-8643 brian.radak_at_anl.gov<mailto:brian.radak_at_anl.gov>
This archive was generated by hypermail 2.1.6 : Sun Dec 31 2017 - 23:21:36 CST