[gmx-users] different output generated by continue and discontinue simulation

Mark Abraham mark.abraham at anu.edu.au
Thu Dec 16 11:34:15 CET 2010



On 12/16/10, Hsin-Lin Chiang  <jiangsl at phys.sinica.edu.tw> wrote:
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Hi,
> 
> 
> 
> My gromacs version is 4.0.5.
> 
> I used grompp to generate tpr for 1ps simulation in this way
> 
> grompp -f md1.mdp -n index.ndx  -p topol.top -c 0ps.gro  -t 0ps.trr -o 1ps.tpr
> 
> mdrun -s 1ps.tpr -e 1ps.edr -o 1ps.trr -g 1ps.log -c 1ps.gro
> 
> Here is my md1.mdp file:
> 
> ;
> 
> title        = ttt
> 
> cpp                 =  /lib/cpp
> 
> constraints         =  hbonds
> 
> ;define              =  -DFLEX_SPC
> 
> integrator          =  md
> 
> emtol               =  100.0
> 
> emstep              =  0.005
> 
> dt                  =  0.002    ; ps !
> 
> nsteps              =  500  ; total 1 ps
> 
> nstcomm             =  500
> 
> nstxout             =  500
> 
> nstvout             =  500
> 
> nstfout             =  500
> 
> nstlog              =  500
> 
> nstenergy           =  500
> 
> nstlist             =  5
> 
> ns_type             =  grid
> 
> rlist               =  1.
> 
> rcoulomb            =  1.
> 
> rvdw                =  1.
> 
> coulombtype         =  PME
> 
> fourierspacing      =  0.12
> 
> pme_order           =  4
> 
> optimize_fft        =  yes
> 
> 
> 
> 
> 
> 

> 
> 
> 
> 
> Tcoupl              =  v-rescale
> 
> tc-grps             =  A-chain B-chain drug SOL  NA+
> 
> 
> 
> 
> 
This is bad - see http://www.gromacs.org/Documentation/Terminology/Thermostats

> 
> 
> 
> 
> 
> ;tau_t               =  0.1  0.1
> 
> tau_t               =  0.2  0.2 0.2 0.2 0.2
> 
> ref_t               =  300.000000 300.000000 300.000000 300.000000 300.000000
> 
> energygrps          =  A-chain B-chain drug SOL  NA+
> 
> Pcoupl              =  berendsen
> 
> Pcoupltype          =  isotropic
> 
> ;tau_p               =  0.1
> 
> tau_p               =  0.25
> 
> compressibility     =  5.4e-5
> 
> ref_p               =  1.0
> 
> gen_vel             =  yes
> 
> gen_temp            =  300.000000
> 
> gen_seed            =  173531
> 
> 
> 
> --------------------------------------------------
> 
> Then I execute below loop in other 59999 times for the totaly 60ns trajectories.
> 
> grompp -f md2.mdp -n index.ndx  -p topol.top -c ${n-1}ps.gro  -t ${n-1}ps.trr -o ${n}ps.tpr
> 
> 
> mdrun -s ${n}ps.tpr -e ${n}ps.edr -o ${n}ps.trr -g ${n}ps.log -c ${n}ps.gro
> 
> 
> 
> 
> 

You are following neither of the approaches recommended here http://www.gromacs.org/Documentation/How-tos/Doing_Restarts


> 
> 
> 
> 
> The file md2.mdp is almost the same as md1.mdp but the parameter gen_vel=no.
> 
> Here ${n-1}ps.gro and ${n-1}.trr mean I use last time frame trajectory to continue.(I didn't really use this commend $(n-1) in bash script, it won't work.)
> 
> 
> 
> I change the other way to run parallel simulation on the same system for double check since above script is difficult to parallelize in public computer.
> 
> The new method is:
> 
> grompp -f md60ns.mdp -n index.ndx  -p topol.top -c 0ps.gro  -t 0ps.trr -o 60ns.tpr
> 
> mpiexec -np 8 mdrun_mpi -s 60ns.tpr -e 60ns.edr -o 60ns.trr -g 60ns.log -c 60ns.gro
> 
> The only different between md1.mdp and md60ns.mdp is nstep=30000000 in md60ns.mdp.
> 
> 
> 
> Theses two data generated by different ways are totaly different.
> 
> Here I mean different is not on the number buy mean the tendency of figure.
> 
> 
> 
> 
> 

See http://www.gromacs.org/Documentation/Terminology/Reproducibility


> 
> 
> 
> 
> 
> In the movie of first method, protein runs violently and go outside the periodic boundary and then split by cubic edge.
> 
> The interaction energy within protein also fluctuate in an unstable way.
> 
> 
> 
> On the contrary, the second method generate a stable movie and fluctuation of interaction energy.
> 
> Theotically these two should be the same, right?
> 
> 
> 
> 
> 

The second is probably happier about not losing the pressure-coupling information every 1ns. However only much
much longer trajectories should show mutual convergence, and the movie is not a reasonable way to look for it.

Mark

> 
> 
> 
> 
> 
> Is anything wrong in my work?
> 
> 
> 
> Sincerely yours,
> 
> Hsin-Lin
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://maillist.sys.kth.se/pipermail/gromacs.org_gmx-users/attachments/20101216/48bc7ea6/attachment.html>


More information about the gromacs.org_gmx-users mailing list