[gmx-users] dihedral restraints with CVS version -- has anyone used dihedral restraints? Do they work?

[David Mobley]

I previously inquired about imposing dihedral restraints in GROMACS.
David responded that I can do this with the CVS version of Gromacs. We
have the CVS version from April 4 and I have been trying to impose
dihedral restraints between a protein and a small molecule (two atoms
in the protein and two in the small molecule, for example).

I know how to implement restraints -- I can do angle and distance
restraints fine. But when I use dihedral restraints, I consistently
get a restraint energy of 0 regardless of how strong I make the
restraints.

Can any developers verify whether the dihedral restraints are known to
actually work? Or have any users on this list actually successfully
used these restraints?

I would appreciate a response. I've e-mailed the list several times
about this and been ignored. I don't know whether it's that people
assume I'm making some stupid mistake and don't want to take the time
to answer my question, or whether someone knows this is a problem and
doesn't want to admit it, or...

I'm about to give up on Gromacs and go back to using AMBER, even
though it's slower, since at least it works. But I figured I should
give this one more try.

Thanks,
David Mobley
UC San Francisco