Re: ABF simulation with extendedLagrangian on keeps crashing

From: Ajasja Ljubetič (ajasja.ljubetic_at_gmail.com)
Date: Thu May 05 2011 - 03:38:16 CDT

Hi,

I gave up on ABF for the time being and ran a metadynamics simulation. But
the simulation still crashed, although I did not see anything in the
trajectory even when writing every step.
The colvar traj is shown
here<http://www.ijs.si/ijs/dept/epr/ajasja/mails/ABF/meta-theta-crash.png>.
I'll try to somehow get rid of the dummy atom and see if it runs better.

Best regards,
Ajasja

colvarsTrajFrequency 1
colvarsRestartFrequency 1000
colvar {
  name theta

  width 5
  lowerboundary 0
  upperboundary 180
  lowerWallConstant 1
  upperWallConstant 1

  angle {

    group1 { dummyAtom (3.283, -0.213, 99.366) } ;
    group2 { atomnumbers 204 } ;#CB
    group3 { atomnumbers 173 } ;#O1

  }
}

colvar {
  name phi

  width 5
  lowerboundary -90
  upperboundary +90
  lowerWallConstant 1
  upperWallConstant 1

  dihedral {

    group1 { atomnumbers 173 } ;#O1
    group2 { atomnumbers 204 } ;#CB
    group3 { dummyAtom (3.283, -0.213, -100.63) } ;
    group4 { dummyAtom (13.283, -0.213, -100.63) } ;
  }
}

metadynamics {
  name CAO_T
  colvars theta
  hillWeight 0.05
  newHillFrequency 100
}

histogram {
   name CBO_TP_hist
   colvars theta phi
}

On Wed, May 4, 2011 at 13:31, Ajasja Ljubetič <ajasja.ljubetic_at_gmail.com>wrote:

> Hi,
>
> I checked the crash in more detail:
> http://www.ijs.si/ijs/dept/epr/ajasja/mails/ABF/bouncing.png
> It looks like the colvar is at the same time attracted to the upper
> boundary and bounces of it.
>
> Here the ABF bias was only active for the theta colvar (the angle colvar)--000e0cd6b2de6efc6c04a2834b1b--

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 05:23:53 CST