PME mesh load

User discussions
1

I am trying to simulate a molecule in water. However, this note shows up. I tried to increase the cut-off (rcoulomb) and the box size but it didn’t work.

NOTE 1 [file min.mdp]:
The optimal PME mesh load for parallel simulations is below 0.5
and for highly parallel simulations between 0.25 and 0.33,
for higher performance, increase the cut-off and the PME grid spacing.

The min.mdp file is:

; RUN CONTROL PARAMETERS =
integrator = steep
emtol = 100
emstep = 0.01

; start time and timestep in ps =
tinit = 0
dt = 0.001
nsteps = 1000000

; number of steps for center of mass motion removal =
comm_mode = none
;nstcomm = 100
;comm_grps = Protein
; OUTPUT CONTROL OPTIONS =
; Output frequency for coords (x), velocities (v) and forces (f) =
nstxout = 100
nstvout = 100
nstfout = 100

; Output frequency for energies to log file and energy file =
nstlog = 100
nstenergy = 100
energygrps = INP

; NEIGHBORSEARCHING PARAMETERS =
; nblist update frequency =
nstlist = 20

; ns algorithm (simple or grid) =
ns_type = grid

; Periodic boundary conditions: xyz or none =
pbc = xyz

; nblist cut-off =
rlist = 2.0
; OPTIONS FOR ELECTROSTATICS AND VDW =
; Method for doing electrostatics =
coulombtype = pme
;coulombtype = cut-off
rcoulomb = 5.0

; Dielectric constant (DC) for cut-off or DC of reaction field =
;epsilon-r = 0.3

; Method for doing Van der Waals =
; cut-off lengths =
cutoff-scheme = verlet
vdw-type = Cut-off
rvdw = 2.0

;OPTIONS FOR TEMPERATURE COUPLING
tcoupl = no
;tc_grps = Protein Other
;tau_t = 0.2 0.2
;ref_t = 300 300
;OPTIONS FOR PRESSURE COUPLING
pcoupl = no
;tau_p = 0.5
;compressibility = 4.5e-05
;ref_p = 1.0
; GENERATE VELOCITIES FOR STARTUP RUN =
;gen_vel = yes
;gen_temp = 300
;gen_seed = -1

; OPTIONS FOR BONDS =
constraints = none
; Type of constraint algorithm =
constraint-algorithm = lincs
lincs_iter = 8
; Do not constrain the start configuration =
;continuation = yes
; Relative tolerance of shake =
;shake-tol = 0.0001
; Highest order in the expansion of the constraint coupling matrix =
lincs-order = 12
; Lincs will write a warning to the stderr if in one step a bond =
; rotates over more degrees than =
lincs-warnangle = 30

2

This is just a note on performance, not correctness. But increasing rcoulomb will not improve performance.

Do you have only a single molecule? In that case use cut-off or reaction-field instead of PME.

3

I have 10 molecules of Anthracene and 450 molecules of water!

I tried to use the coulombtype = cut-off and the warning was that using this could let to artifacts!

4

If you have that many molecules simply use PME and standard settings (for the force field chosen). There is no point in increasing cut-offs.

5

Thank you Hess!