You print out a restart file with your simulation. Simply replace the file name you're using for the coordinate flag (-c) with your restart file (.rst or whatever you put for the restart flag (-r) in your old simulation), and your simulation will continue off from whenever that restart file was made.
Ex: $ sander -O -i 03_Heat2.in -o 03_Heat2.out -p DOPC_128.prmtop -c 02_Heat.rst -r 03_Heat.rst -ref 02_Heat.rst -x 03_Heat.nc
My coordinate file in this case is 02_Heat.rst and I'm making a restart file with the name 03_Heat.rst. You're .in file you can set the frequency at which you output a set of a restart coordinates.
Be sure to set ntx=5 and irest=1 in your input (.in) file so that it reads the velocities from your restart file. And you can use ntwr to set how often the restart file is written.
Charlie
________________________________________
From: Sreemol G [sreemolinfo.gmail.com]
Sent: Tuesday, April 05, 2016 10:42 PM
To: AMBER Mailing List
Subject: Re: [AMBER] General question
Thank you. That works. Is there any powercut command for amber to restart
the simulation from the step where power cut occurs as we give in gromacs
(mdrun -v -s 10.tpr -e 10.edr -g 10.log -x 10.xtc -cpi 10.cpt -c 10.gro
-cpo 10.cpt -append & )
On Wed, Apr 6, 2016 at 10:13 AM, Charles Lin <clin92.ucsd.edu> wrote:
> mpirun -np 4 $AMBERHOME/bin/sander.MPI -O -i blah.in -o blah.out ...
>
> The variable after -np allows you to choose the number of processors you
> wish to run for your mpi task.
>
> You may need to set MPI_HOME to your MPI directory and export the path.
>
> Charlie
>
> ________________________________________
> From: Sreemol G [sreemolinfo.gmail.com]
> Sent: Tuesday, April 05, 2016 9:31 PM
> To: david.case.rutgers.edu; AMBER Mailing List
> Subject: Re: [AMBER] General question
>
> Thanks for your suggestions. I have configured using -mpi gnu and installed
> ambertools15. But I dont have Amber14 installed. I'm running sander.MPI
> even now it is running in single node (| Running AMBER/MPI version on 1
> nodes) . What could I do to run in 4 nodes??
>
> On Mon, Apr 4, 2016 at 6:31 PM, David A Case <david.case.rutgers.edu>
> wrote:
>
> > On Mon, Apr 04, 2016, Sreemol G wrote:
> >
> > > configure: error: could not find mpi library for --enable-mpi"
> > >
> > > my system has mpi/openmpi-x86_64 installed in it.
> >
> > It's almost sure that the system mpi is using a different compiler than
> the
> > one you specified in the configure step (which was gnu 5.3). You can
> check
> > this by typing
> >
> > mpicc -show
> >
> > (or something similar). The simplest way to get a compatible mpi is
> > to cd to $AMBERHOME/AmberTools/src and run "configure_mpich" -- see the
> > comments at the top of this file for instructions.
> >
> > An alternative (depending on your experience and the type of computer you
> > are
> > running on) is to figure out which compiler your system mpi is using, and
> > configure with that option.
> >
> > Also, you will almost certainly get better performance by running
> pmemd.MPI
> > in place of sander.MPI (if you have Amber14 installed).
> >
> > ...dac
> >
> >
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
> >
>
>
>
> --
> With kind regards,
> G. Sreemol
> M.Tech (Computational Biology)
> Anna university
> Chennai.
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Tue Apr 05 2016 - 23:00:04 PDT
