On Thu, Jan 17, 2013 at 12:01 PM, DEBOSTUTI GHOSHDASTIDAR <
debostutighosh.gmail.com> wrote:
> Dear Prof. Case,
>
> Thank you for your reply. I equilibrated the system without restraint
> at 298K for 1ns. Since the solvent is ionic in nature, I expect long
> range electrostatics and have used a cut off of 15.
I think you're misunderstanding what 'cut' means in Ewald (or particle-mesh
Ewald) simulations. The cutoff does _not_ truncate electrostatic
interactions -- those are calculated out to infinity (look up the Ewald
summation method -- the particle-mesh part allows you to use a DFFT instead
of a normal Fourier transform (to drastically speed things up). The cut
divides the part of the electrostatic sum that is done in direct space
(slow) versus reciprocal space (fast). It also establishes the cutoff for
Lennard-Jones interactions, which decay as 1/r^6 and are therefore
short-range. For more info, see Dave's email.
> We have used the
> same molecule and its force field parameters for simulation of other
> systems (not with DNA though). What should I do now? Should I continue
> with dt=0.001 for longer. If the force field is the problem, should'nt
> the system have a problem with dt=0.001 also?
No. Force field errors and integration errors are unrelated (even if they
seem related). Even an incorrect force field will conserve energy
(obviously as long as the forces are conservative, which Amber's are).
Integration errors, caused by the fact that we have to use a finitely
large (rather than infinitely small) time step as well as limited-precision
coordinates and velocities, cause the problems you're seeing. This is
caused when the forces are 'too large' for your time step -- there are many
reasons this may be true.
HTH,
Jason
> Thanks for your help
>
> On 1/17/13, David A Case <case.biomaps.rutgers.edu> wrote:
> > On Thu, Jan 17, 2013, DEBOSTUTI GHOSHDASTIDAR wrote:
> >>
> >> I am simulating DNA in a solvent. After thorough minimization, I
> >> heated up the system to 298K keeping the DNA restrained and thereafter
> >> performed NPT at 298K keeping the restraints on. However, the system
> >> showed a SHAKE error saying "SHAKE cannot be accomplished". So I read
> >> out the last few frames just before the error and using VMD saw that a
> >> C-H bond of one molecule of the solvent is stretched too much. When I
> >> repeated the simulation, the error occured again, this time for a
> >> different solent molecule and a N-H bond. During this entire procedure
> >> i was using dt=0.002.
> >
> > I doubt that very many people have experience with simulation of DNA in
> > solvents other than water. There may be large forces on the solvent
> > molecules, and it is certainly possible that there are problems with
> > whatever
> > force field parameters you are using for the solvent. Or, you may still
> > need
> > to equilibrate for longer (you don't say how long the equilibration
> was...)
> > It's not too uncommon to need to use dt=0.001 for the first stages of a
> > simulation.
> >
> >>
> >> Here's my input for md
> >> &cntrl
> >> imin = 0, ntx = 7, irest = 1,
> >> ntpr = 1000, ntwx = 1000, ntwe = 0,
> >> ntc = 2, ntf = 2,iwrap = 1,nstlim = 500000, dt =0.002,
> >> tempi=298.0, temp0 = 298.0, ntt =1, tautp =2.0,
> >> cut = 15.0, ntb = 2,ntp = 1,taup = 2.0,pres0 = 1.0
> > ^^^^^^^^^^^^
> >
> > This is way too big....consider using a value in the range of 8-10 Ang.
> >
> > ...dac
> >
> >
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
> >
>
>
> --
> Debostuti Ghosh Dastidar
> PhD Scholar
> Computational Biophysics Lab
> Department of Biotechnology
> Indian Institute of Technology Madras
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
--
Jason M. Swails
Quantum Theory Project,
University of Florida
Ph.D. Candidate
352-392-4032
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Jan 17 2013 - 12:00:02 PST
