Improper Angles

Gerard Kleijwegt gerard at
Sat Nov 18 14:25:23 EST 1995

In article <30ACEBFB.167E at>, Gu Peihua <pku at> writes:
|> To Whomever Can Help,
|> I'm sorta new to X-PLOR and molecular dynamics in general.  I've been
|> reading the manual and find that I don't understand what an "improper"
|> angle is.  How is that different from a dihedral angle?
|> If anyone can help, I'd greatly appreciate it.
|> Peihua Ku
|> UCLA Physiology

one for the FAQ ?

a dihedral is a physical torsion angle, e.g.:

  CG-CD1-()-CE1-CZ in Phe and Tyr

here the torsion is around the CD1-CE1 bond; to
keep the aromatic rings flat you need to restrain
this torsion to zero (plus several others)

an improper is a "virtual" torsion angle defined
so as to keep an atom and its bonded neighbours
in a certain configuration

ex. 1 - a chiral carbon with four different
  neighbours (e.g., CA in all amino acids except Gly)
  CA-N-C-CB is an improper torsion (about the
  non-existent N-C bond); it's value is ~+34 degrees
  for an L amino acid, -34 for a D amino acid

ex. 2 - a carboxylate carbon has to be flat (e.g.,
  in aspartate); the improper torsion CG-CB-OD1-OD2
  should be zero

ex. 3 - an sp2-hybr. C has to be flat, e.g. C=C-C
   in retinol; again an improper defined with the
   sp2 carbon first, followed by its three neighbours
   should be zero

note: if the improper is zero, the order of the three
neighbours doesn't matter; if it's non-zero there are
two choices with a sign (and chirality) difference

i tend to use high weights for my impropers and fixed
dihedrals (750 when using the engh & huber force field)


More information about the X-plor mailing list