Equilibrating for 10 fs - mdp file

User discussions
1

GROMACS version: 2023.1-conda_forge
Martini: 3

Hello,
I am fairly new to modelling membranes and I am trying to simulate an asymmetric membrane. I minimized it and now when I am trying to equilibrate it, it gives, I get the following fatal error. I will paste both my code and my error.
I would appreciate if somebody with experience in this field could help me modify my mdp parameters as well. According to the papers I have been reading, I think I should use the Berendsen for semi-isotropic.
Below is my code and the error:

code and the error:

gmx grompp -f equilibrate_10fs.mdp -c EColi-LIPA-CG_EM.gro -p Ecoli-LIPA-AMP.top -o Ecoli-LIPA-AMP-EQ10.tpr -n EColi-LIPA-CG_EM.ndx

Ignoring obsolete mdp entry ‘title’
Ignoring obsolete mdp entry ‘ns_type’

WARNING 1 [file equilibrate_10fs.mdp]:
The Berendsen barostat does not generate any strictly correct ensemble,
and should not be used for new production simulations (in our opinion).
For isotropic scaling we would recommend the C-rescale barostat that also
ensures fast relaxation without oscillations, and for anisotropic scaling
you likely want to use the Parrinello-Rahman barostat.

Setting the LD random seed to -1219498291

Generated 844 of the 356590 non-bonded parameter combinations

Excluding 1 bonded neighbours molecule type 'states the lipid and water and other components of my lipid bilayer"

Setting gen_seed to -671627314

Velocities were taken from a Maxwell distribution at 310 K

Cleaning up constraints and constant bonded interactions with virtual sites
Number of degrees of freedom in T-Coupling group Membrane is 77896.46
Number of degrees of freedom in T-Coupling group Solvent is 73816.54

Determining Verlet buffer for a tolerance of 0.005 kJ/mol/ps at 310 K

Calculated rlist for 1x1 atom pair-list as 1.181 nm, buffer size 0.081 nm

Set rlist, assuming 4x4 atom pair-list, to 1.133 nm, buffer size 0.033 nm

Note that mdrun will redetermine rlist based on the actual pair-list setup

This run will generate roughly 541 Mb of data

There was 1 WARNING

Program: gmx grompp, version 2023.1-conda_forge
Source file: src/gromacs/gmxpreprocess/grompp.cpp (line 2691)

Fatal error:
Too many warnings (1).
If you are sure all warnings are harmless, use the -maxwarn option.

my mdp file includes the following:

title = Martini

;#define -DPOSRES

integrator = md
dt = 0.01
nsteps = 10000000
nstcomm = 100

nstxout = 0
nstvout = 0
nstfout = 0
nstlog = 1000
nstenergy = 1000
nstxout-compressed = 5000
compressed-x-precision = 100
compressed-x-grps =
energygrps = Membrane Solvent

cutoff-scheme = Verlet
nstlist = 20
ns_type = grid
pbc = xyz
verlet-buffer-tolerance = 0.005

coulombtype = reaction-field
rcoulomb = 1.1
epsilon_r = 15 ;2.5 (with polarizable water)
epsilon_rf = 0
vdw_type = cutoff
vdw-modifier = Potential-shift-verlet
rvdw = 1.1

tcoupl = v-rescale
tc-grps = Membrane Solvent
tau_t = 1.0 1.0
ref_t = 310 310

pcoupl = Berendsen
pcoupltype = semiisotropic
tau_p = 12
compressibility = 3e-4 3e-4 3e-4
ref_p = 1.0 1.0 1.0
refcoord-scaling = all

gen_vel = yes
gen_temp = 310
gen_seed = -1

constraints = none
constraint_algorithm = Lincs
lincs_order = 8
lincs_warnangle = 90
lincs_iter = 2

2

What you get is a warning, you should read it and make sure that you understand what the problem is. The recommendation is to switch to a better barostat (c-rescale), but for equilibrating the system, it should be OK with Berendsen (even if there is no reason to use it except to exactly reproduce old simulations).

If you are sure that you can proceed with your current settings, in spite of the warning you can use the -maxwarn 1 option when running gmx grompp.

3

Thank you for your help.
My problem is that somebody else has ran this simulation in my lab before using -maxwarn and the production files were no good. So we decided to not use -maxwarn and find a solution for the equilibration step so that it runs by itself.
Would there be any reason that it is not running and keeps giving error even though it is semi-isotropic (with Berendsen)?

Thank you

4

Berendsen is not good for production runs, since it does not generate a proper ensemble. But if you’re doing equilibration, it should not matter. For production, I’d certainly recommend switching to c-rescale - it works for semi-isotropic pressure coupling as well. Except for more accurate reproducibility there’s no reason to use Berendsen anymore, not even for equilibration.

But I don’t think this will address the problems with the previous production simulations, as it is difficult to spot that an incorrect ensemble was used.