[gmx-users] Umbrella sampling's equilibration runs

Justin Lemkul jalemkul at vt.edu
Tue Apr 30 17:09:08 CEST 2013

On 4/30/13 10:58 AM, James Starlight wrote:
> Dear Gromacs users!
> I have a question about umbrella sampling simulation based on the Justin's
> tutorial.
> According to the tutorial after definition of the set of conformers
> extracted from the pulled trajectory I should run N equilibrating
> simulations and N productions runs. In the tutorial I've found that all
> equilibrations run in the NPT ensemble. In my case I have membrane receptor
> for each conformer extracted from the pulling trajectory I want to run 20
> umbrella's simulations with different starting velocities in each case.
> Should I re-equilibrate each conformer in the nvt+npt runs (re-assigning
> velocities in the nvt run) or the velocities might be re-assigned in the
> npt equilibrations ? What the time-prolongation of each equilibrations
> should be for each conformer in case of membrane protein simulation?

If you re-assign velocities to start NPT after NVT, what was the point of NVT? 
You destroy the previously established state.  Initialize velocities at the 
start of NVT, then preserve the ensemble information when moving to NPT, like 
any other simulation.

The time frame is something you must decide based on your knowledge and 
observations of your system.  There is no definitive answer.



Justin A. Lemkul, Ph.D.
Research Scientist
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080


More information about the gromacs.org_gmx-users mailing list