Pressure bolwing up and down

User discussions
1

Hello, I’m running a simulation of glucose oxidase in water with ions (Cl and Na). I have problems running the position restraint with the pressure as it never get to 1 bar but gets to very high or very low levels.
I’m using GROMACS 2022, the box dimensions are 8.73600 9.04600 8.73600.
Here is the mdp file:
define = -DPOSRES
constraints = h-bonds
pbc = xyz
integrator = md
dt = 0.002 ; ps !
nsteps = 100000 ; total 200 ps.

; Output control
nstxout = 10000 ; every 20 ps
nstvout = 10000
nstfout = 0
nstlog = 5000
nstenergy = 5000
nstcalcenergy = 10
nstxout-compressed = 100000
;compressed-x-precision = 10000
;compressed-x-grps = System

; Center of Mass motion
comm_mode = Linear
nstcomm = 10
comm_grps = Prot_glycan_FAD Water_and_ions
; Neighborsearching
cutoff-scheme = Verlet
nstlist = 10 ; 20 fs
;ns_type = grid
;rlist = 1.2 ; short-range neighborlist cutoff in nm
;rcoulomb = 1.2
coulombtype = PME
fourierspacing = 0.12
vdw-modifier = Potential-switch
rvdw_switch = 0.8

; Temperature coupling
Tcoupl = V-rescale
tc-grps = Prot_glycan_FAD Water_and_ions
tau_t = 0.1 0.1
ref_t = 300 300

; Energy monitoring
energygrps = Prot_glycan_FAD Water_and_ions

; P coupling is on
pcoupl = C-rescale
pcoupltype = isotropic
tau_p = 4.0
ref_p = 1.0
compressibility = 4.5e-5
refcoord_scaling = all

; Dispersion correction
DispCorr = no

; Velocity generation
Continuation = no
gen_vel = yes
gen_temp = 300
gen_seed = -1

resulting pressures are
Pressure (bar)
1.75038e+01 -4.29765e+01 -1.51139e+01
-4.29816e+01 3.94336e+01 7.51545e+00
-1.51090e+01 7.51555e+00 -7.71275e+01

Thanks in advance to anyone that will anser.

2

When you say “it never get to 1 bar”, do you mean that you have not reached one bar after 200 ps equilibration?

If you plot the pressure during those 200 ps, what does it look like? What is the starting pressure and what is the end pressure, i.e., average over the last 50 ps?

200 ps is very short for a reasonably large system. Have you tried extending it to 5 or 10 ns?

Also, keep in mind that pressure can fluctuate quite a lot. You should be aiming at an average pressure of close to 1 bar - when the pressure is stable.

3

When I say it never gets to 1 bar I mean it surpass the pressure of 1 bar by some folds both negatively and positively. The average of those 200ps is far off -31bar.

I’ll try by extending the relaxation time, thank you. May I already ask, if extending the relaxation time wouldn’t work, what could be the most probable error?
Thank you again

4

Since a large part of those 200ps is before the equilibrium is reached, you should not use all of that simulation when calculating the pressure. Use the -b option of gmx energy. You are interested in the average pressure when you have reached a plateau, hopefully the last 50ps might give some indication. If the pressure is still drifting at that point you almost certainly want to equilibrate longer.

In general, there should not be any problem reaching a correct average pressure if you simulate long enough. However, you will see that the pressure is fluctuating a lot during a simulation. The larger the simulation system the smaller the pressure fluctuations. Your system is fairly large, so I don’t think the standard deviation of the pressure (at equilibrium) will be extremely high, but still perhaps 15-50 bar.