Re: 2.7b metadynamics ::alpha module :: folding of a single helix in water

From: Ali Emileh (ali.emileh_at_gmail.com)
Date: Tue Jun 09 2009 - 11:52:19 CDT

Giacomo,

Thank you for the reply.
I removed the "conflicting" hillWidth line and now use these settings.

    width 0.04

    lowerBoundary 0.3
    upperBoundary 1.0

    hillWeight 0.04

but it's behaving the same way. Of note, are two warning lines in the log
file :

colvars: Warning: colvar interval ( 3.00000000000000e-01 -
1.00000000000000e+00) is not commensurate to its width (
4.00000000000000e-02).
..
..
..
colvars: Setting a new lower boundary for colvar "a1", at -0.38.
colvars: Expanding the grids for the metadynamics bias "metadynamics1".
colvars: Warning: colvar interval (-3.80000000000000e-01 -
1.00000000000000e+00) is not commensurate to its width (
4.00000000000000e-02).

So it's ignoring my boundaries. I can't understand why.

I have decreased the newhillfrequency parameter to 10, hoping that it will
squeeze the
protein into SOMETHING!, in hope of some CHANGE. Right now, metadynamics
seems
to have no visible effect. The forces output in the .traj file have
increased to +/- 800-1000,and their absolute value seems to increase
linearly with time. Does it mean its applying
the forces? Do I need more time? By applying it to this small system, I just
want to make
sure that metadynamics works so I can move on and apply a similar
configuration to my
main system.

Any thoughts about the above warnings and also what I should do to 'see'
something?

Thanks a lot,
Ali

On Fri, Jun 5, 2009 at 11:53 AM, Giacomo Fiorin <gfiorin_at_seas.upenn.edu>wrote:

> Hi Ali, the grid resolution is set by the "width" of a1: use that, and
> only that, to change the resolution of the PMF.
>
> hillWidth, instead, is only a smoothing parameter because it is
> multiplied to the number above as the documentation says, and should
> never be set smaller than 1. If you set it to 0.02, all of the hills
> become 0.02*0.04 = 0.0008 wide: way too narrow to have an effect.
>
> Giacomo
>
> ---- ----
> Giacomo Fiorin
> Center for Molecular Modeling at
> University of Pennsylvania
> 231 S 34th Street, Philadelphia, PA 19104-6323
> phone: (+1)-215-573-4773
> fax: (+1)-215-573-6233
> mobile: (+1)-267-324-7676
> mail: giacomo.fiorin_<at>_gmail.com
> web: http://www.cmm.upenn.edu/
> ---- ----
>
>
>
> On Wed, Jun 3, 2009 at 1:39 AM, Ali Emileh<ali.emileh_at_gmail.com> wrote:
> > Hi all,
> >
> > I have been trying to use metadynamics to fold a simple alpha helix (16
> > residues long)
> > without any success. Contacting the module authors, I refined my Gaussian
> > sizes (which
> > were too broad) but still can't get it to work and the alpha colvars I
> have
> > defined doesn't
> > look to be changing much other than what you would expect from a normal
> MD.
> > During a
> > 10 ns run, it barely changed from the strating value of ~0.44 to ~0.5
> toward
> > the end. I had
> > set a lowerboundary of 0.5 for the colvars but I'm guessing since it
> started
> > lower than that,
> > it might be ignoring the boundary? In the starting conformation, the
> helix
> > is half folded and
> > I have used TMD before to easily fold it in the context of its parent
> > molecule (this is an
> > excised helix I'm using to learn about the metadynamics module). This is
> how
> > my
> > COLVARS.CONF file looks like:
> > ==============================================
> > colvarsTrajFrequency 1000
> >
> > colvar {
> > name a1
> > width 0.04
> > outputAppliedForce on
> > lowerBoundary 0.4
> > upperBoundary 1.0
> >
> > alpha {
> > residueRange 100-116
> > psfSegID G
> > }
> > }
> >
> > metadynamics {
> > colvars a1
> > hillWeight 0.05
> > hillWidth 0.02
> > dumpFreeEnergyFile yes
> > }
> > ==============================================
> >
> > I have tried lower boundaries of 0.3 and 0.5 and also hillweights and
> widths
> > of (0.4,0.2,0.1) (0.02,0.05,0.1),
> > respectively. None worked.
> >
> > I would highly appreciate ant input.
> >
> > Thank you.
> >
> > Ali
> >
>

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:52:54 CST