GROMACS version 2021:
GROMACS modification: No
Hello,
My name is Matteo and I’m currently learning GROMACS so I have no experience at all.
After NVT equilibration I have always a negative pressure (-300 bar).
I want to have a positive pressure of about 560 bar without using barostat but only varying box size.
Can someone help me ?
I’m using this nvt script.
http://www.mdtutorials.com/gmx/lysozyme/06_equil.html
You have to equilibrate the system under the desired conditions and ensemble, so in your case if you want an NVT ensemble but with fixed pressure, you have to equilibrate under NPT at that pressure and once you achieve it, switch back to NVT.
Ok thank you.
So the first equilibration is NPT and the second is NVT ?
The NVT phase wouldn’t necessarily be considered equilibration, but yes, the order I’m suggesting is NPT then NVT.
Ok thank you.
GROMACS version 2021
GROMACS modification: No
Hello,
My name is Jackson and I’ve encountered a similar issue. I do indeed start with a 5 ns NPT equilibration run at 1 bar where I find an average pressure +/- RMSD of (-0.85 +/- 134.6) bar, and a box size from 2-5 ns of (7.29 +/- 0.002) nm. I was taught that prior to NVT I should adjust the box size to the average value over NPT and repeat an energy minimization run. I did both these steps, but my next NVT run has a pressure of around (-214.6 +/- 135.4) bar where it has stayed for 300 ns now.
Is this behavior expected, or could it indicate that I need to run the first NPT for longer? It does look like 5ns NPT was plenty looking at energy metrics (potential, temperature, pressure).
Thanks so much!
It seems like you’ve got large volumes of vacuum in your system and that the particles are not close enough to attract each other. So, the system is not exploding, but slowly expanding.
From what starting point did you start your NPT simulations? The most common order (but not the only) would be: energy minimization (EM) → NVT equilibration → NPT equilibration.
If you have large volumes of vacuum in the system after the NVT equilibration you should try a much smaller simulation box or try:
EM → NVT → NPT at very high pressure → NPT
This order could be changed, but the take-home message is that you need to get your box smaller before doing NPT equilibrations at normal (atmospheric?) pressure.
Thanks so much your response!
So does a negative pressure indicate the system is expanding? I would have though the opposite, i.e. a negative pressure indicates the box wants to collapse. Also if there are large volumes of vacuum, would that be something I could see by making some sort of histogram of water positions? It’s difficult to tell by eye looking at VMD visualizations.
In any case, I actually began my first NPT simulation directly after energy minimization, so:
EM → NPT → Adjust Box → EM → NVT. I can definitely try beginning with EM → NVT prior to my first NPT step, as well as a high pressure step, and that does make sense that I need the box to be smaller to raise the pressure, which I suppose is also in agreement with the fact that decreasing pressure (eventually negative) implies an expanding box.
Thanks again, and I’ll try the extra equilibration steps to see if this helps.
Best,
Jackson
You are correct that a negative pressure (too large box) should be compensated by the barostat (by shrinking the box), but sometimes the expanding driving force is too large. It’s possible that a very low tau-p (with a first-order barostat) would be better than applying a high pressure (as I recommended above). But in most cases, either should work (I think). If you cannot see any void spaces with, e.g., VMD, it’s possible that the issue is something else.
I don’t really see the purposes of the “EM → NPT → Adjust Box → EM → NVT” setup (unless you want your production simulation in the NVT ensemble, of course), but there are possibly systems where that is good or necessary. In most cases I think EM → NVT → NPT is a safer recommendation. Hopefully you won’t even need any high pressure stages.