[gmx-users] Block averaging

[Justin Lemkul]

On 6/22/16 5:02 AM, sun wrote:
> Hello users and experts I have completed a 200 ns protein ligand simulation
> using GROMOS 43a1 and Gromacs v 5.0. I expected to observe a conformational
> change in protein in the presence of ligand and the results are as expected
> and correlates to previous references as well. So, shall i believe that
> simulation is converged or there is need to do block averaging over time
> intervals? If block averaging is required, the parameters like RMSD and
> propensities for secondary structure are sufficient to conclude that
> simulation is converged? Or could anyone please tell me a suitable procedure
> for making blocks and how to average those. With Regards Suniba
>

Just because a simulation produces expected results does not mean it is 
converged.  You must show that any quantities of interest from which you draw 
your conclusions, are not systematically varying with time.

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==================================================