AW: Namd list: "Size cell" for restart

From: Norman Geist (norman.geist_at_uni-greifswald.de)
Date: Thu Aug 30 2012 - 01:59:07 CDT

Hi,

 

please check the coor file you use with vmd(as pdb if not binary) if it is
really wrapped (water diffused??). If not, and you don't use a xsc file,
namd will use the bounding box of the system to guess the dimensions, what
would give such large dimensions due to the diffused waters. If you was
using a xsc, check the xsc file (text editor) to see if the box informations
are written there that large. If they are written there that large,
something went wrong with the previous simulation. Alsio just post the xsc
file and the namd config file of the previous run, and the one that is not
working now.

 

Norman Geist.

 

Von: owner-namd-l_at_ks.uiuc.edu [mailto:owner-namd-l_at_ks.uiuc.edu] Im Auftrag
von Suranga Edirisinghe
Gesendet: Dienstag, 28. August 2012 20:45
An: Axel Kohlmeyer
Cc: Mariana Graterol; norman.geist_at_uni-greifswald.de; jpotoffx_at_gmail.com;
namd-l_at_ks.uiuc.edu; namd_at_ks.uiuc.edu
Betreff: Re: namd-l: Namd list: "Size cell" for restart

 

Hi
I would let namd to use default parameters as mush as possible to start
with. Then slowly make changes as system runs smoothly.

Thanks
Suranga

On Tue, Aug 28, 2012 at 9:01 AM, Axel Kohlmeyer <akohlmey_at_gmail.com> wrote:

On Tue, Aug 28, 2012 at 3:07 PM, Mariana Graterol
<marianagraterol_at_gmail.com> wrote:
> Hello and thanks to all,
>
> After NPT run I can't restart the calculation, cause the "size cell" is
too
> big now. I know that in a molecular dynamics with pressure and temperature
> controls, volume is not constant... but, my initial cell was 120 x 120 x
120
> Angs., and then it increases until 1100 x1100 x 1100!

which means that something went horribly wrong.

you may want to run with fixed volume for a bit
before you enable npt. also, you may need to
run a minimization to reduce the potential energy.

please check out the namd tutorials and their discussion
of what kind of protocol to follow to set up a periodic
system and why. this must not be followed blindly,
but with adaptations for your specific system.

axel.

>
> The structure file looks like the cubic box is losted, that is, all water
> molecules are scattered around the protein a very large distance.
> I have wrap all on, and langevin pressure control in my configuration
file;
> starting with a minimized structure and previously heated slowly until 300
> K.
>
> Is it really my cell size? Why the "box" image is lost? Is it something
> about coordinates or a mistake in setting npt conditions?
>
> Thank you for any comment that help me...
>
> ******
> # Input files
> coordinates C:\\Users\\pc\\Documents\\test.pdb
> structure C:\\Users\\pc\\Documents\\test.psf
> parameters parm.prm
> paraTypeCharmm on
>
> # Output files
> binaryoutput yes
> outputname C:\\Users\\pc\\Documents\\test
> restartfreq 1000
> restartsave yes
> binaryrestart yes
> DCDfile C:\\Users\\pc\\Documents\\test.dcd
> DCDfreq 1000
> DCDUnitCell no
> outputEnergies 1000
> mergeCrossterms yes
> outputMomenta 0
> outputPressure 0
> outputTiming 10000
>
> # Timestep parameters
> numsteps 200000
> timestep 1.000000
> firsttimestep 0
> stepspercycle 20
>
> # Simulation space partitioning
> cutoff 12.000000
> switching on
> switchdist 8.000000
> limitdist 0.000000
> pairlistdist 14.000000
> splitPatch hydrogen
> hgroupCutoff 2.500000
> margin 0.000000
> pairlistMinProcs 1
> pairlistsPerCycle 2
> outputPairlists 0
> pairlistShrink 0.010000
> pairlistGrow 0.010000
> pairlistTrigger 0.300000
>
> # Basic dynamics
> exclude scaled1-4
> temperature 300.000000
> COMmotion no
> zeroMomentum no
> dielectric 1.000000
> nonbondedScaling 1.000000
> 1-4scaling 1.000000
> vdwGeometricSigma no
> seed 12345
> rigidBonds none
>
> # PME parameters
> PME on
> PMETolerance 1.000000e-06
> PMEInterpOrder 4
> PMEGridSpacing 1.000000
> PMEGridSizeX 120
> PMEGridSizeY 150
> PMEGridSizeZ 144
> FFTWEstimate no
> FFTWUseWisdom yes
>
> # Full direct parameters
> FullDirect no
>
> # Multiple timestep parameters
> fullElectFrequency 4
> MTSAlgorithm impulse
> longSplitting c1
>
> # Harmonic constraints
> constraints on
> consexp 2
> conskcol B
> constraintScaling 1.000000
> conskfile C:\\Users\\pc\\Documents\\test.pdb
> consref C:\\Users\\pc\\Documents\\test.pdb
>
> # Periodic boundary conditions
> cellBasisVector1 117.002419 0.000000 0.000000
> cellBasisVector2 0.000000 145.766808 0.000000
> cellBasisVector3 0.000000 0.000000 143.076828
> cellOrigin -0.000001 -0.000001 0.0000006
> XSTfreq 1000
> wrapWater off
> wrapAll on
> wrapNearest off
>
> # Langevin dynamics
> langevin on
> langevinTemp 300.000000
> langevinHydrogen off
> langevinDamping 5.000000
>
> # Interactive molecular dynamics
> IMDon on
> IMDport 3000
> IMDfreq 20
> IMDwait off
> IMDignore off
>
> # Constant Pressure Control
>
> useGroupPressure no
> useFlexibleCell no
> useConstantArea no
> langevinPiston on
> langevinPistonTarget 1.01325
> langevinPistonPeriod 100
> langevinPistonDecay 50
> langevinPistonTemp 300
>
>
>
> --
> * mari *

--
Dr. Axel Kohlmeyer  akohlmey_at_gmail.com  http://goo.gl/1wk0
International Centre for Theoretical Physics, Trieste. Italy.
 

This archive was generated by hypermail 2.1.6 : Mon Dec 31 2012 - 23:21:59 CST