[gmx-users] NPT simulation stage

[Justin Lemkul]

Please keep the discussion on the list.  I can see from the other posts that 
providing this information to everyone likely would have been helpful.

On 4/28/14, 2:44 AM, Mahboobeh Eslami wrote:
> dear justin
> I sent you the actual results for the repeated three times NPT stage:
>
> Energy                      Average   Err.Est.       RMSD  Tot-Drift
> -------------------------------------------------------------------------------
> Pressure                   -1.86895        1.7    74.4741    8.46672  (bar)
>
> Energy                      Average   Err.Est.       RMSD  Tot-Drift
> -------------------------------------------------------------------------------
> Pressure                   -1.01376        3.1    76.8484   0.464786  (bar)
>
> Energy                      Average   Err.Est.       RMSD  Tot-Drift
> -------------------------------------------------------------------------------
> Pressure                   0.251751        3.7    80.9027   -1.89681  (bar)
>
>   if i change tau_p for the second NPT stage to 6 i will get following result:
>
> Energy                      Average   Err.Est.       RMSD Tot-Drift
> -------------------------------------------------------------------------------
> Pressure                    1.20898        1.9    78.1862   0.931217  (bar)
>
> Energy                      Average   Err.Est.       RMSD  Tot-Drift
> -------------------------------------------------------------------------------
> Pressure                  -0.514578        1.8    78.6336 -2.57617  (bar)
>
> Energy                      Average   Err.Est.       RMSD  Tot-Drift
> -------------------------------------------------------------------------------
> Pressure                    5.67055        1.9    80.5453    2.32409  (bar)
>

Increasing tau_p relaxes the stringency of the barostat; I would expect the 
results to be "worse" when doing so.

The bigger point here is that none of this looks anomalous to me.  You're doing 
short NPT (100 ps), during which the pressure is unlikely to relax fully.  More 
importantly, look at the actual numbers: -2 ± 75, -1 ± 77, 0.2 ± 81.  Are any of 
these significantly different from the target value of 1 or do they differ 
significantly from one another?  I'd say no.  Pressure is ill-defined and 
subject to very large fluctuations.  I wouldn't call any of this cause for concern.

-Justin

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

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA 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
http://mackerell.umaryland.edu/~jalemkul

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