[AMBER] Query regarding restarting the replica exchange molecular dynamics simulation

From: Munazzah Fatima Ansari via AMBER <amber.ambermd.org>
Date: Wed, 22 Mar 2023 17:51:02 +0530

Hello Experts,

I was running the REMD simulation. But due to some reasons it stopped.
Now I want to restart the simulation again. For restarting the
simulation I made the new input file same as the previous one, only
difference is that in place of ig = RANDOM_NUMBER I used ig = -1 for all
12 replicas in remd1.mdin file.
 My remd1.mdin.001 file is given below:

 &cntrl

  imin = 0,
   irest = 1,
                                                                        
                            
   ntx = 5,

   ntt = 3,
   gamma_ln = 2.0,
   temp0 = 300.00,
   ig = -1,

   ntc = 2,
   ntf = 2,
   ntb = 0,

  igb = 5,
   cut = 999.0,
   rgbmax = 999.0,

   ntpr = 5000,
   ntwx = 5000,
   ntwr = 5000,

   nmropt = 1,
   nscm = 500,

   numexchg = 10000,
   nstlim = 25000,
   dt = 0.002,
  &end
 /
 &wt TYPE='END'
 /
DISANG=/home/munazzah/Remd/Test_for_restart/Remd_imp_droso/5.Production/1u

Then in remd1.groupfile I use -A flag in order to append output to
existing output files.
The command in groupfile for 12 replicas given below :

-A -rem 1 rem.log -i remd1.mdin.001 -o remd.mdout.001 -c remd.rst.001 -r
remd.rst.001 -x remd.mdcrd.001 -inf remd.mdinfo.001 -p 1urr.prmtop

Then I run this job with the command:

nohup mpirun -np 12 sander.MPI -ng 12 -groupfile remd1.groupfile &

But it is overwriting in output file, I am attaching the output file
named as remd.mdout.001.

When I use the same command with pmemd.cuda.MPI in place of sander.MPI
then it returns,

Running multipmemd version of pmemd Amber22
    Total processors = 12
    Number of groups = 12

  unknown flag: -A
                                                                        
                                                          

  usage: pmemd [-O] -i mdin -o mdout -p prmtop -c inpcrd -r restrt
              [-ref refc -x mdcrd -v mdvel -frc mdfrc -e mden -inf
mdinfo -l logfile]
              [-ng numgroups -groupfile groupfile -rem remd_method]
              [-rremd rremd_type -reservoir reservoir.nc]
              [-hybridsolvent_remd_traj hybridsolvent_remd_traj_name]
              [-amd amdlog_name -gamd gamdlog_name -scaledMD
scaledMDlog_name -suffix output_files_suffix]
              [-sf sfout_name]

  unknown flag: -A
                                                                        
                                                          

  unknown flag: -A
                                                                        
                                                          

  unknown flag: -A
                                                                        
                                                      

  unknown flag: -A
.............

Unknown flag for all 12 replica. I am not understanding why simulation
is not run with -A flag for appending the output and why simulation is
not running with pmemd.cuda.MPI and why it is overwriting the output
file as I use -A flag for append?

Thanking you

Munazzah Fatima Ansari

_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber

Received on Wed Mar 22 2023 - 05:30:03 PDT
Custom Search