I am trying to restrain torsions and bond angles to do
minimizations using sander_classic (do not want Cartesian coordinate
restraints). I have tried to use the restrainTorsion command in xleap,
which for all I can tell is working fine (desc [unit] tells me that the
restraints I specified has been applied.) But, after I run sander and
look at my coordinates I don't see the effect of the restraints.
It is possible that you are the first person to use actually
restrainTorsion and have discovered a bug in the way leap
writes the prmtop. This is because leap came into use around
the time that nmr restraints were implemented, and everyone
started using these. I think it would be reasonable to remove
the restrain* commands from leap at this point.
Hopefully your questions about the nmr restraints can be answered
and this will solve your problem.
Bill Ross
Received on Wed Jul 11 2001 - 15:40:55 PDT
