Re: Water Box Shape Changing during NVT Simulation!

From: Aditya Ranganathan (aditya.sia_at_gmail.com)
Date: Tue Mar 09 2010 - 04:12:29 CST

By steps I mean, each part of the simulation, like minimization, heating,
equilibration and the production run. Assume that the initial
CellBasisVector values of the solvated system was 108,160,108. I used these
values to set up the heating step, then again checked for the minmax and
center of the structure obtained after the slow heating and used these
values for the next step. So eventually, the CellBasisVector values for the
final production run were far different from the ones which I obtained and
used for the first minimization/heating steps.

Is this the reason for this problem? I`ve attached a file of an image with
all the periodic cell images turned on (the number of images on vmd set to
1). The image has a view of all the periodic cell images (+X.+Y,+Z,-X,-Y,-Z)
 turned on and has screenshots from different angles. Please let me know if
this is an incorrect or inappropriate or has bubbles and vacuum.

Thanks,
Aditya

On Tue, Mar 9, 2010 at 3:08 PM, Axel Kohlmeyer <akohlmey_at_gmail.com> wrote:

> aditya,
>
> On Tue, Mar 9, 2010 at 1:03 AM, Aditya Ranganathan <aditya.sia_at_gmail.com>
> wrote:
> >
> > Thanks to all for the suggestions.
> > However my question still remains and I`m sorry that I`m sounding so
> naive.
> > Since I`ve been recalculating periodic cell values at each step and not
> > keeping it the same as what it was in the first step, I was wondering if
>
> i don't quite understand what you mean by this. you said initially, that
> you
> did an NVT ensemble simulation. that means that in the simulation the
> cell shape and size does not change. what do your mean by "step"
> and "recalculate" in this context?
>
> > this is the reason for the change of shape as I have increased the
> periodic
> > cell values and thus allowing the system to sample the additional space
> and
>
> if you increase the cell size for a dense periodic system, you will
> add vacuum. this is a rather unrealistic model. specifically, if your
> initial conditions are assembled from "pieces" (i.e. protein, bilayer,
> solvent)..
>
> > thus the change of shape. Also, will this actually have any adverse
> effect
> > on the way the protein actually is behaving and thus the accuracy of the
> > whole simulation?
> > As I said, the peptides are still well inside the solvent and not peeping
> > out.
>
> from what you describe, you have one or multiple droplets and not a
> dense system. this has severe consequences, since the water vaccuum
> interface will induce a surface dipole potential due to the orientational
> preference of water molecules in the interface. this will make the peptides
> see a very different environment from a regular bulk system.
>
> > When I turn on the periodic cell images option in VMD, do all the
> periodic
> > cells have to be close to each other without space in between? Please
> help
>
> yes.
>
> > me out with the same. I`ll post a screenshot with all the periodic cells
> > displayed on VMD.
>
> please do so. a picture says much more than many words in this case.
>
> cheers,
> axel.
> >
> > Thanks
> > Aditya
> >
> > When I switch on all the periodic cell images
> >
> > On Tue, Mar 9, 2010 at 2:35 AM, Jeffrey J. Potoff
> > <jpotoff_at_chem1.eng.wayne.edu> wrote:
> >>
> >> Ronald Salesky wrote:
> >>>
> >>> Hi Aditya,
> >>>
> >>>>> My question also is whether the volume can still remain the same even
> >>>>> with the shape of the water box changing to cylindrical? Does an NVT
> >>>>> simulation place a restriction on the shape or just the total volume
> of the
> >>>>> water box. If its just the volume, cant the water molecules just
> arrange and
> >>>>> rearrange themselves without the volume changing but just the
> shape?<<
> >>>
> >>> My understanding is the PBC (periodic boundary condition) you have
> placed
> >>> on your system sets a periodic cell with the x,y,z boundaries you have
> >>> specified, which means the system is modeled as if it is really a
> periodic
> >>> array of the same protein/solvent system like in a crystal. Since the
> >>> solvent/protein system wants to converge to its minimum energy, the
> solvent
> >>> shape within the cell will change. If you were to just model a single
> cube
> >>> of water with no solute molecule it would seek its minimum energy state
> >>> which would be spherical. I am learning NAMD/VMD as you are, so those
> more
> >>> experienced may disagree with some of these points.
> >>>
> >> Just a point of clarification. In a canonical (NVT) MD simulation, one
> is
> >> minimizing the free energy of the system, not the energy.
> >>
> >> --
> >> ======================================================================
> >> Jeffrey J. Potoff jpotoff_at_chem1.eng.wayne.edu
> >> Associate Professor Wayne State University
> >>
> >> Department of Chemical Engineering and Materials Science
> >> 5050 Anthony Wayne Dr Phone:(313)577-9357
> >> Detroit, MI 48202 Fax: (313)578-5815
> >> http://potoff1.eng.wayne.edu
> >> ======================================================================
> >>
> >
> >
>
>
>
> --
> Dr. Axel Kohlmeyer akohlmey_at_gmail.com
> http://sites.google.com/site/akohlmey/
>
> Institute for Computational Molecular Science
> Temple University, Philadelphia PA, USA.
>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:55:33 CST