AMBER Archive (2002)

Subject: multiple improper torsion terms

From: Thijs Beuming (Thijs.Beuming_at_mssm.edu)
Date: Fri Jan 11 2002 - 21:37:58 CST

Hello,

I am deriving a forcefield for a molecule in the excited state. Since
the molecule is non-planar in the excited state I am trying to define an
improper torsion that would bias the forcefield in such a way to keep it
non-planar during minimisation/dynamics. The difference between quantum
mechanical and AMBER energy (without the imp. torsion term) can only be
reproduced with 3 or more cosine expansions. However, unlike for normal
dihedrals, it seems that multiple terms give problems for improper
torsions in AMBER.
So
 CB-C -N*-C 1.98 -144.52 1.
gives the correct energy but
CB-C -N*-C 3.86 31.34 -3.
CB-C -N*-C -12.39 201.31 -2.
CB-C -N*-CT 1.98 -144.52 1.
doesn't give the correct energy....

Does anybody have a solution ?

Thijs Beuming