Re: Energy problem after restart

From: Fangqiang Zhu (fzhu_at_ks.uiuc.edu)
Date: Mon Nov 24 2003 - 15:15:56 CST

This is really strange. I tried to restart a simulation with very similar
configurations as yours, but got exactly the same electrostatic energy
as the original run. In other word, I can't reproduce your problem.
If you can send me the standard output of your restart simulation,
maybe I could help out further...

Zhu

Alessandro Pedretti wrote:

> Thank you for the answer !
> I removed the line:
>
> extendedsystem $MolFile.restart.xsc
>
> from the resatart file, but the problem is still present: the electrostatic
> energy is undersetimated. After the restart, it seems that the dielectric
> constants is changed from 1 to 1.17, but it's funny !
>
> Alessandro
>
> ----- Original Message -----
> From: "Fangqiang Zhu" <fzhu_at_ks.uiuc.edu>
> To: "Alessandro Pedretti" <alessandro.pedretti_at_unimi.it>
> Cc: <namd-l_at_ks.uiuc.edu>
> Sent: Friday, November 21, 2003 12:15 AM
> Subject: Re: namd-l: Energy problem after restart
>
> > Hi,
> >
> > I think your simulation was intended to be of vacuum type (i.e., not
> > periodic).
> > Therefore, in your restart simulation, you shouldn't specify
> "extendedsystem".
> >
> > When that command is present, NAMD will do the simulation with periodic
> > boundary condition, and that's why you see different electrostatic energy.
> >
> > Zhu
> >
> >
> > Alessandro Pedretti wrote:
> >
> > > Hi all !
> > > I restarted my simple dynamics but I found that the energy of the last
> step
> > > of the first simulation and the starting step of the second one is
> > > different.
> > >
> > > Last step of the first dynamics:
> > >
> > > TIMING: 860000 CPU: 354917, 0.589419/step Wall: 354917, 0.589419/step,
> > > 22.9218 hours remaining, 0 kB of memory in use.
> > > ENERGY: 860000 3709.6508 5116.2602 2992.5010
> > > 6.9421 -21150.7474 -9771.7463 0.2589 0.0000
> > > 12444.9404 -6491.9403
> 144.6908 -6383.8469 -6383.9937
> > > 144.7454
> > >
> > > WRITING EXTENDED SYSTEM TO RESTART FILE AT STEP 860000
> > > WRITING COORDINATES TO DCD FILE AT STEP 860000
> > > WRITING COORDINATES TO RESTART FILE AT STEP 860000
> > > FINISHED WRITING RESTART COORDINATES
> > > WRITING VELOCITIES TO RESTART FILE AT STEP 860000
> > > FINISHED WRITING RESTART VELOCITIES
> > > LDB: LOAD: AVG 54.73 MAX 61.5 MSGS: TOTAL 54 MAXC 36 MAXP 1 None
> > > LDB: LOAD: AVG 54.73 MAX 55.67 MSGS: TOTAL 54 MAXC 36 MAXP 1 Refine
> > >
> > > Step "0" of the second simulation:
> > >
> > > ENERGY: 860000 3709.6487 5116.2603 2992.5012
> > > 6.9423 -18110.9999 -9771.7462 0.0000 0.0000
> > > 12444.8849 -3452.5088
> 144.6901 -3343.9650 -3343.9650
> > > 144.6901
> > >
> > > The electrostatic energy is different (-18110.9999 instead
> of -21150.7474)
> > > and I don't understand what is wrong !
> > > My system configuration is:
> > > - Dual AMD Athlon MP 2200
> > > - Windows 2000
> > > - NAMD 2.5 for Windows
> > > At the end of this e-mail, you can find the input files for the two
> > > simulations.
> > > Thank you
> > >
> > > Alessandro Pedretti
> > >
> > > --
> > > Dr. Alessandro Pedretti
> > > Ist. di Chimica Farmaceutica e Tossicologica
> > > Viale Abruzzi, 42
> > > I-20131 Milano (Italy)
> > > Tel. +39 02 503 17522
> > > Fax +39 02 503 17565
> > > E-Mail: alessandro.pedretti_at_unimi.it
> > > WWW: http://www.ddl.unimi.it
> > >
> > > ================================================================
> > > This is the first aborted dynamics file input:
> > >
> > > # Protocol params
> > > numsteps 1000000
> > >
> > > # Initial config
> > > coordinates a1aa_adre_mem_aq_min.pdb
> > > temperature 300K
> > > seed 12345
> > >
> > > # Output params
> > > outputname a1aa_adre_mem_aq_dyn
> > > outputEnergies 5000
> > > outputtiming 5000
> > > binaryoutput yes
> > > DCDFreq 5000
> > > restartFreq 5000
> > >
> > > # Integrator params
> > > timestep 1.0
> > > nonbondedFreq 2
> > > fullElectFrequency 4
> > >
> > > # Force field params
> > > structure a1aa_adre_mem_aq_min.psf
> > > paraTypeCharmm on
> > > parameters ../LIP/par_all22_prot_lipid.inp
> > > exclude scaled1-4
> > > 1-4scaling 0.4
> > > switching on
> > > switchdist 8.0
> > > cutoff 12.0
> > > pairlistdist 14.0
> > > margin 0.0
> > > stepspercycle 20
> > >
> > > # Cylindrical PBC
> > > cylindricalBC on
> > > cylindricalBCCenter 0.0 0.0 0.0
> > > cylindricalBCAxis y
> > > cylindricalBCr1 42.00
> > > cylindricalBCl1 62.00
> > > cylindricalBCk1 2
> > > ================================================================
> > > ... and this is the restart file:
> > >
> > > set MolFile a1aa_adre_mem_aq_dyn
> > > set ResSuffix 2
> > > set ParFile ../LIP/par_all22_prot_lipid.inp
> > >
> > > set StartStep 860000
> > > set Steps 1500000
> > > set Refresh 5000
> > >
> > > # Please remember that "Steps" must be the sum of steps of the first
> > > dynamics
> > > # and the steps of the second one.
> > >
> > > # End of user parameters
> > >
> > > numsteps $Steps
> > > seed 12345
> > > coordinates a1aa_adre_mem_aq_dyn.restart.pdb
> > > bincoordinates $MolFile.restart.coor
> > > binvelocities $MolFile.restart.vel
> > > extendedsystem $MolFile.restart.xsc
> > >
> > > outputname $MolFile$ResSuffix
> > > outputEnergies $Refresh
> > > binaryoutput yes
> > > DCDFreq $Refresh
> > > restartFreq $Refresh
> > >
> > > firsttimestep $StartStep
> > > timestep 1.0
> > > nonbondedFreq 2
> > > fullElectFrequency 4
> > >
> > > structure a1aa_adre_mem_aq_min.psf
> > > paraTypeCharmm on
> > > parameters $ParFile
> > > exclude scaled1-4
> > > 1-4scaling 1.0
> > > switching on
> > > switchdist 8.0
> > > cutoff 12.0
> > > pairlistdist 14.0
> > > margin 0.0
> > > stepspercycle 20
> > >
> > > # Cylindrical PBC
> > > cylindricalBC on
> > > cylindricalBCCenter 0.0 0.0 0.0
> > > cylindricalBCAxis y
> > > cylindricalBCr1 42.00
> > > cylindricalBCl1 62.00
> > > cylindricalBCk1 2
> > > ================================================================
> >

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 05:17:56 CST