[gmx-users] Disk Space
Justin A. Lemkul
jalemkul at vt.edu
Tue Jun 15 01:16:41 CEST 2010
teklebrh at ualberta.ca wrote:
> Dear Chris and Justin,
>
> I just read the manual and it stated that I need a .trr file if I want
> to extend the simulation ( from 5ns -20ns).
>
> My question is:
>
> For the first 5ns I want to use berendsen thermostat and barostat so
> that my system will relax faster compared to N-H and P-R. Then I will
> switch on to the Nose-Hover and Parrinello-Rahman ( this will give me a
> correct description for the kinetic energy information.
>
> Therefore, I need both coordinate and velocity files then ( Gromacs 3.3.3).
>
>>>> nstxout = 50000
>>>> nstvout = 50000
>>>> nstfout = 0
>
>>>> ; Output frequency for energies to log file and energy file
>>>> nstlog = 2000
>>>> nstenergy = 1000
>>>> ; Output frequency and precision for xtc file
>>>> nstxtcout = 2000
>>>> xtc-precision = 1000
>
> rather than set the values of nstxout and nstvout , if I set them to
> 50000, I could get the .trr file. But for my entire analysis I will
> depend on the .xtc file.
>
> step 1
>
> after 5ns of my simulation( Both berendsen thermostat and barostat. I
> have to use still grompp to generate a new .tpr file by modifying the
> .mdp file ( t_ coupling,p_couple and the tinit). Once done this I run
> mdrun.
>
> step 2
>
> use the following procedure then after
>
> http://www.gromacs.org/Documentation/How-tos/Extending_Simulations
>
> tpbconv -s previous.tpr -f previous.trr -e previous.edr -o next.tpr
> -extend timetoextend
>
> Am I correct? and also is the output file seems reasonable.
>
That's certainly a reasonable way to extend a simulation. With 3.3.3, you need
full-precision coordinates and velocities. Is there any reason you want to use
software that is several years old? The upgrades in the 4.0.x series are quite
nice, and make extending simulations far easier, requiring only a checkpoint
file. This (I assume) is why Chris said that the .trr file is becoming obsolete.
> I am trying to balance both the disk space issue and what output I need
> to continue further.
>
If you only need the .trr file to continue your simulation, you only need the
very last frame, so there is really no reason to save high-precision coordinates
every few ps for 5 ns. Your .trr file may exceed several GB in size, all for no
reason!
-Justin
> Thank you chris and justine.
>
> Rob
>
>
>
>
> Quoting "Justin A. Lemkul" <jalemkul at vt.edu>:
>
>>
>>
>> teklebrh at ualberta.ca wrote:
>>> Dear Chris,
>>>
>>> Thank you very much!
>>>
>>> I am planning to set the output control option like this!
>>>
>>> tinit = 0
>>> nstxout = 2000
>>> nstvout = 0
>>> nstfout = 0
>>> nstlog = 2000
>>> nstenergy = 1000
>>>
>>> I want to set the velocity and force option to zero and save the
>>> coordinate and energy every 1000 or 2000 steps ( 2ps or 4ps). By
>>> doing this am I missing anything.
>>>
>>
>> Unless you need full-precision coordinates every few ps, you can use
>> nstxtcout = 2000 and set nstxout = 0. A .xtc file is substantially
>> smaller than a .trr file.
>>
>> -Justin
>>
>>> regards,
>>>
>>> Rob
>>>
>>>
>>>
>>> Quoting chris.neale at utoronto.ca:
>>>
>>>> Dear Rob:
>>>>
>>>> To answer your question: just set up your run with nsteps for 5 ns,
>>>> run mdrun, and then use tpbconv -until. A simpler option is to set
>>>> it up to run 20 ns, but use -maxh to limit it to an amount of
>>>> runtime that you pre-calculate to be similar to 5 ns of simulation.
>>>>
>>>> HOWEVER:
>>>>
>>>> You almost surely don't need to save this so often: nstxout = 100
>>>> (also nstvout, nstfout).
>>>>
>>>> And you may not even need to save this so often: nstxtcout = 100
>>>> (also nstenergy).
>>>>
>>>> I had no idea that there was an option nstcheckpoint = 100 as I use
>>>> mdrun -cpt 6 to specify a .cpt file every 6 hours, but there is no
>>>> need to write a .cpt every 100 steps.
>>>>
>>>>
>>>>
>>>> Dear Gromacs users,
>>>>
>>>> I am running an MD simulation but it looks to me that I am running out
>>>> of disk space.
>>>>
>>>> This is my .mdp file
>>>>
>>>> ; 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
>>>> ; Output frequency and precision for xtc file
>>>> nstxtcout = 100
>>>> xtc-precision = 100
>>>> nstcheckpoint = 100 ; checkpointing helps you continue
>>>> after crashe. This checkpint file is saved every 1ns
>>>>
>>>>
>>>> Please tell me which files i need to set zero.
>>>>
>>>> I am planning to run my simulation for 20ns but I want it to run every
>>>> five nanoseconds until i see an aggregations in my system. So how can
>>>> I re-start my simulation every 5ns.
>>>>
>>>>
>>>> Thank you for your help
>>>>
>>>> rob
>>>>
>>>>
>>>>
>>>> --
>>>> gmx-users mailing list gmx-users at gromacs.org
>>>> http://lists.gromacs.org/mailman/listinfo/gmx-users
>>>> Please search the archive at http://www.gromacs.org/search before
>>>> posting!
>>>> Please don't post (un)subscribe requests to the list. Use thewww
>>>> interface or send it to gmx-users-request at gromacs.org.
>>>> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>>>>
>>>>
>>>
>>
>> --
>> ========================================
>>
>> Justin A. Lemkul
>> Ph.D. Candidate
>> ICTAS Doctoral Scholar
>> MILES-IGERT Trainee
>> Department of Biochemistry
>> Virginia Tech
>> Blacksburg, VA
>> jalemkul[at]vt.edu | (540) 231-9080
>> http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin
>>
>> ========================================
>> --
>> gmx-users mailing list gmx-users at gromacs.org
>> http://lists.gromacs.org/mailman/listinfo/gmx-users
>> Please search the archive at http://www.gromacs.org/search before
>> posting!
>> Please don't post (un)subscribe requests to the list. Use the www
>> interface or send it to gmx-users-request at gromacs.org.
>> Can't post? Read http://www.gromacs.org/mailing_lists/users.php
>>
>>
>
>
--
========================================
Justin A. Lemkul
Ph.D. Candidate
ICTAS Doctoral Scholar
MILES-IGERT Trainee
Department of Biochemistry
Virginia Tech
Blacksburg, VA
jalemkul[at]vt.edu | (540) 231-9080
http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin
========================================
More information about the gromacs.org_gmx-users
mailing list