[gmx-users] NPT or NVT

[Justin Lemkul]

On 1/2/14, 5:34 AM, Riccardo Concu wrote:
> Dear all, I’ve a doubt/question. I ran a long simulation of my system (100ns)
> at 298K. Now I want to see the effect of the T on my system and I want to run
> a 20ns simulation at 350K and 400K, now my question is: Do I have to run this
> simulation under the NPT condition or under the NVT (using an annealing for
> example?). I’m trying to run the simulation at 350K using the final gro of my
> simulation but the system is very unstable and finishes due to too many lincs
> warning.

You will need to thoroughly re-equilibrate the system under whatever new 
conditions you impose.  Annealing may be useful.  Higher temperatures also may 
require a shorter time step to be stable.  As for ensemble, NVT will be more 
stable, but you need to decide what to use based on what is relevant, not what 
may happen to "just work."  You can make NPT work, but proper preparation and 
input settings will be required.

-Justin

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

Justin A. Lemkul, Ph.D.
Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 601
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441

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