Dear Amber user,
I was trying to run an explicit solvent cpHMD simulation on a membrane protein system with amber18. After equilibrating the system in NVT ensemble for 200ps, I tried to move the system to NPT ensemble with MC barostat. However, MC barostat will cause the system to stuck indefinitely (~12 hours) in the first step but Berendsen will keep the system going normally.
MC barostat do work properly in non-MPI mode in a local machine on CPU. After equilibrating with MC barostat for 200ps, the system has been moved to the cluster to run in MPI mode again in a 16 core CPU node but it still stuck indefinitely, where the last few lines of the output file is:
4. RESULTS
--------------------------------------------------------------------------------
---------------------------------------------------
APPROXIMATING switch and d/dx switch using CUBIC SPLINE INTERPOLATION
using 5000.0 points per unit in tabled values
TESTING RELATIVE ERROR over r ranging from 0.0 to cutoff
| CHECK switch(x): max rel err = 0.3316E-14 at 2.351880
| CHECK d/dx switch(x): max rel err = 0.8066E-11 at 2.747240
---------------------------------------------------
|---------------------------------------------------
| APPROXIMATING direct energy using CUBIC SPLINE INTERPOLATION
| with 50.0 points per unit in tabled values
| Relative Error Limit not exceeded for r .gt. 2.34
| APPROXIMATING direct force using CUBIC SPLINE INTERPOLATION
| with 50.0 points per unit in tabled values
| Relative Error Limit not exceeded for r .gt. 2.81
|---------------------------------------------------
| MC Barostat: Decreasing size of volume moves
| MC Barostat: Decreasing size of volume moves
The input file is:
Relax in NPT
&cntrl
imin = 0, ! No minimization.
ig = -1, ! pseudo-random number generator
ntx = 5, ! Coordinates and velocities will be read
irest = 1, ! Restart the simulation
ntpr = 2500, ! Frequency of printing of the energy and its components.
ntwx = 2500, ! Every ntwx steps, the coordinates will be written to the mdcrd file.
ntwr = 10000,! Every ntwr steps during dynamics, the “restrt” file will be written.
ntf = 2,
ntc = 2, ! bonds involving hydrogen are constrained
ntb = 2, ! constant pressure
ntp = 2, ! 1:isotropic, 2:anisotropic, 3:semiisotropic
barostat = 2, ! Monte Carlo barostat
mcbarint = 50, ! Number of steps between the Monte Carlo barostat
pres0 = 1.0, ! Reference pressure
cut = 10.0, ! Non-bond cut off
nstlim = 2500000,! Perform MD for 500,000 steps = 1 ns
nscm = 1250, ! removal of COM
dt = 0.002, ! timestep
tempi = 300.0, ! Initial temperature.
temp0 = 300.0, ! Reference temperature at which the system is to be kept,
ntt = 3, ! Langevin dynamics
gamma_ln = 2.0, ! Collision frequency
ntr = 1, ! position restraint
restraintmask = '(:1-207)&(!.H=)',
restraint_wt = 25.0, !The weight (in kcal/mol−Å^2) for the positional restraints.
icnstph=2 ! cpHMD in explicit solvent
ntcnstph=100 ! Attempt protonation every 100 steps
ntrelax=100 ! 200 fs relaxed solvent distribution
solvph=7.0 ! pH of the Golgi pH
saltcon=0.1 ! Physiological salt concentration
/
Kind regards,
Zhiyi Wu
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Sat May 11 2019 - 14:30:02 PDT
