Re: COM restraint in NAMD

From: JC Gumbart (
Date: Wed Aug 22 2007 - 11:49:13 CDT

If you are content to only restrain in one direction (x, y, or z
only), the standard SMD commands in NAMD with a velocity of zero will
effectively do what you want. I'm not so familiar with FreeEnergy
restraints, but you're right that your second option is far from

On Aug 22, 2007, at 10:55 AM, Victor Ovchinnikov wrote:

> Dear all,
> I am performing Steered MD (using the TclForces interface) on a
> solvated
> protein and would like to restrain the center of mass of the
> protein to
> keep it from drifting.
> Can any of you suggest an efficient way of doing this?
> Two ways came to my mind:
> 1) Use FreeEnergy restrains -- specify a positional restraint on
> the COM
> of all atoms; the problem is that (correct me if I messed up) if an
> atom
> participates in FE restraints, it's coordinates are unavailable to
> `calcforces'; in order not to include SMD atoms in the FE restraint
> atoms, I'd have to come up with an elaborate selection string in
> `urestraint { ... '
> 2) Modify the TclForces script to add a COM restraining force in
> addition to the steering force;
> My question here is whether doing this would slow the code
> substantially, since I'd have to calculate the COM at every timestep &
> loop over all of the atoms in `calcforces'
> Let me know if you have suggestions/comments.
> Cheers,
> Victor

This archive was generated by hypermail 2.1.6 : Wed Feb 29 2012 - 15:45:08 CST