[gmx-users] average pressure of a system

[Justin Lemkul]

On 9/10/13 9:12 PM, Dwey wrote:
> Hi All.
>
>     I work on a simulation of a membrane protein in a model membrane
> (DPPC) with a total of 26,859 atoms.  In a step of NPT equilibration,
> my mdp file is used for a 100 ps NPT and it has a reference pressure
> of 1 bar.
>
>   At the end of simulation,  I obtained a result of the average
> pressure of the system ( ~ -0.90 bar)  like
>
> Energy                      Average   Err.Est.       RMSD  Tot-Drift
> -------------------------------------------------------------------------------
> Pressure                    -0.9049        2.6    186.304    13.8503  (bar)
>
>
> I am wondering if  I should keep running this NPT process until the
> average pressure of the system reaches ~ 1.0 bar.
>
> If so, how long ( how many steps) ?
>
> Or should I modify mdp file  ? Can anyone provide some suggestions ?
>

If you want commentary on your .mdp file, you need to provide the actual .mdp file.

>
> Likewise, at the end of Production MD for 1 ns, I obtained  average
> pressure (~2.23 bar)  of the system like
>
> Energy                      Average   Err.Est.       RMSD  Tot-Drift
> -------------------------------------------------------------------------------
> Pressure                    2.23349        2.1     164.97    10.9381  (bar)
>
> Should I run a longer simulation until the average pressure reaches ~
> 1.0 bar,  although average energy, average temperature (323 K),
> average density (1022 kg/m^3) are already at  the desired values ?
>
> How should I do to stabilize average pressure  at a desired value (~1 bar) ?
>

None of this looks abnormal to me at all.  This topic is discussed very 
frequently on the list, so please refer to the archive for extensive discussion 
about the theory of pressure calculations and what the output means.  Also have 
a look at http://www.gromacs.org/Documentation/Terminology/Pressure and ask 
yourself whether 2 ± 165 is significantly different from the target value of 1. 
  Also realize the 1 ns is extremely short for a membrane system, which normally 
take tens of ns to equilibrate.

-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

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