[gmx-users] umbrella sampling tutorial

[Justin Lemkul]

On 4/7/15 7:11 AM, Ming Tang wrote:
> Hi, Justin
>
> I am working with a triple helix using umbrella sampling method. When
> generating configurations, I pulled 1ns (md_pull.mdp). Then I got 15
> configurations using 0.03 spacing. They are conf0.gro conf5.gro conf10.gro
> conf15.gro conf17.gro conf18.gro conf30.gro conf31.gro conf39.gro conf59.gro
> conf87.gro conf161.gro conf233.gro conf360.gro conf488.gro. Most of the
> configurations are within 0-100 frames. Does this affect the umbrella
> sampling simulations? Is there any pulling time requirement when generating
> the configurations?  I was supposed to have a try and find the result, but
> found that the calculation is time-consuming. So I decided to ask you
> finally.
>

The physics of every system is different.  It's hard to generalize.  If you have 
configurations that allow for reasonable window spacing, the quality of the end 
result depends on the sampling and convergence of the umbrella sampling itself.

-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

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