<br /><br /><span>On 12/16/10, <b class="name">Hsin-Lin Chiang </b> <jiangsl@phys.sinica.edu.tw> wrote:</span><blockquote cite="mid:20101216082255.M46640@phys.sinica.edu.tw" class="iwcQuote" style="border-left: 1px solid rgb(0, 0, 255); padding-left: 13px; margin-left: 0pt;" type="cite"><div class="mimepart text html"><span><p>
<meta content="text/html; charset=big5" http-equiv="Content-Type" />
<meta content="OPENWEBMAIL" name="GENERATOR" />
<table><tbody><tr><td bgcolor="#ffffff"><p>
Hi,
<br />
<br />My gromacs version is 4.0.5.
<br />I used grompp to generate tpr for 1ps simulation in this way
<br />grompp -f md1.mdp -n index.ndx -p topol.top -c 0ps.gro -t 0ps.trr -o 1ps.tpr
<br />mdrun -s 1ps.tpr -e 1ps.edr -o 1ps.trr -g 1ps.log -c 1ps.gro
<br />Here is my md1.mdp file:
<br />;
<br />title = ttt
<br />cpp = /lib/cpp
<br />constraints = hbonds
<br />;define = -DFLEX_SPC
<br />integrator = md
<br />emtol = 100.0
<br />emstep = 0.005
<br />dt = 0.002 ; ps !
<br />nsteps = 500 ; total 1 ps
<br />nstcomm = 500
<br />nstxout = 500
<br />nstvout = 500
<br />nstfout = 500
<br />nstlog = 500
<br />nstenergy = 500
<br />nstlist = 5
<br />ns_type = grid
<br />rlist = 1.
<br />rcoulomb = 1.
<br />rvdw = 1.
<br />coulombtype = PME
<br />fourierspacing = 0.12
<br />pme_order = 4
<br />optimize_fft = yes
<br /></p></td></tr></tbody></table></p></span></div></blockquote><span></span><blockquote cite="mid:20101216082255.M46640@phys.sinica.edu.tw" class="iwcQuote" style="border-left: 1px solid rgb(0, 0, 255); padding-left: 13px; margin-left: 0pt;" type="cite"><div class="mimepart text html"><span><p><table><tbody><tr><td bgcolor="#ffffff"><p>Tcoupl = v-rescale
<br />tc-grps = A-chain B-chain drug SOL NA+
</p></td></tr></tbody></table></p></span></div></blockquote>This is bad - see http://www.gromacs.org/Documentation/Terminology/Thermostats<br /><blockquote cite="mid:20101216082255.M46640@phys.sinica.edu.tw" class="iwcQuote" style="border-left: 1px solid rgb(0, 0, 255); padding-left: 13px; margin-left: 0pt;" type="cite"><div class="mimepart text html"><span><p><table><tbody><tr><td bgcolor="#ffffff"><p><br />;tau_t = 0.1 0.1
<br />tau_t = 0.2 0.2 0.2 0.2 0.2
<br />ref_t = 300.000000 300.000000 300.000000 300.000000 300.000000
<br />energygrps = A-chain B-chain drug SOL NA+
<br />Pcoupl = berendsen
<br />Pcoupltype = isotropic
<br />;tau_p = 0.1
<br />tau_p = 0.25
<br />compressibility = 5.4e-5
<br />ref_p = 1.0
<br />gen_vel = yes
<br />gen_temp = 300.000000
<br />gen_seed = 173531
<br />
<br />--------------------------------------------------
<br />Then I execute below loop in other 59999 times for the totaly 60ns trajectories.
<br />grompp -f md2.mdp -n index.ndx -p topol.top -c ${n-1}ps.gro -t ${n-1}ps.trr -o ${n}ps.tpr
<br />
mdrun -s ${n}ps.tpr -e ${n}ps.edr -o ${n}ps.trr -g ${n}ps.log -c ${n}ps.gro
</p></td></tr></tbody></table></p></span></div></blockquote><br />You are following neither of the approaches recommended here http://www.gromacs.org/Documentation/How-tos/Doing_Restarts<br /><br /><blockquote cite="mid:20101216082255.M46640@phys.sinica.edu.tw" class="iwcQuote" style="border-left: 1px solid rgb(0, 0, 255); padding-left: 13px; margin-left: 0pt;" type="cite"><div class="mimepart text html"><span><p><table><tbody><tr><td bgcolor="#ffffff"><p>The file md2.mdp is almost the same as md1.mdp but the parameter gen_vel=no.
<br />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.)
<br />
<br />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.
<br />The new method is:
<br />grompp -f md60ns.mdp -n index.ndx -p topol.top -c 0ps.gro -t 0ps.trr -o 60ns.tpr
<br />mpiexec -np 8 mdrun_mpi -s 60ns.tpr -e 60ns.edr -o 60ns.trr -g 60ns.log -c 60ns.gro
<br />The only different between md1.mdp and md60ns.mdp is nstep=30000000 in md60ns.mdp.
<br />
<br />Theses two data generated by different ways are totaly different.
<br />Here I mean different is not on the number buy mean the tendency of figure.
</p></td></tr></tbody></table></p></span></div></blockquote><br />See http://www.gromacs.org/Documentation/Terminology/Reproducibility<br /><br /><blockquote cite="mid:20101216082255.M46640@phys.sinica.edu.tw" class="iwcQuote" style="border-left: 1px solid rgb(0, 0, 255); padding-left: 13px; margin-left: 0pt;" type="cite"><div class="mimepart text html"><span><p><table><tbody><tr><td bgcolor="#ffffff"><p><br />In the movie of first method, protein runs violently and go outside the periodic boundary and then split by cubic edge.
<br />The interaction energy within protein also fluctuate in an unstable way.
<br />
<br />On the contrary, the second method generate a stable movie and fluctuation of interaction energy.
<br />Theotically these two should be the same, right?
</p></td></tr></tbody></table></p></span></div></blockquote><br />The second is probably happier about not losing the pressure-coupling information every 1ns. However only much<br />much longer trajectories should show mutual convergence, and the movie is not a reasonable way to look for it.<br /><br />Mark<br /><blockquote cite="mid:20101216082255.M46640@phys.sinica.edu.tw" class="iwcQuote" style="border-left: 1px solid rgb(0, 0, 255); padding-left: 13px; margin-left: 0pt;" type="cite"><div class="mimepart text html"><span><p><table><tbody><tr><td bgcolor="#ffffff"><p><br />Is anything wrong in my work?
<br />
<br />Sincerely yours,
<br />Hsin-Lin
<br />
</p></td></tr></tbody></table>
</p></span>
</div></blockquote>