AMBER: Assignment of parameters, missing parameters

From: Julien Michel <>
Date: Mon, 22 Mar 2004 15:47:15 +0000

Dear Amber users,

        I would like to report that Leap can assigns torsions incorrectly when
generating a prmtop file. The problem seems to be caused by atom typing.
How is it possible to differentiate the following torsion ca-ca-ca-ca in a
phenyl ring and ca-ca-ca-ca in a biphenyl(GAFF atom types) ? As it currently
stand, with AMBER7, leap does not make a difference between the torsion
linking the two phenyl rings and the'usual' torsion and incorrectly assign a
wrong torsion. The OPLS forcefield has two different torsions in this case (the
torsion around the bond linking the two phenyls rings has a smaller amplitude,
faciliting ring rotation), while I can only locate the torsion X-ca-ca-X in
gaff or(X -CA-CA-X in AMBER). Even if I were to generate parameters for the
missing torsion, how would I force leap to recognize it ? Manually editing the
prmtop file, even with scripts, is certainly not practical.
        The best solution might be to introduce a new atom type for
something like "aromatic sp2 carbon linked to another atom not in the ring
which has more than one bond". I do not have AMBER8, and I would like to know
if leap/GAFF2 can handle this case ?

Julien Michel
University of Southampton
Department of Chemistry
Dr J. W. Essex group 
The AMBER Mail Reflector
To post, send mail to
To unsubscribe, send "unsubscribe amber" to
Received on Mon Mar 22 2004 - 16:53:00 PST
Custom Search