[gmx-users] dihedral potential OPLSAA force field

[David Huang]

Hi

I've converted the OPLS-AA torsional potential parameters for the
alkane C-C-C-C (CT-CT-CT-CT in the gromacs parameter file notation),
C-C-C-H (CT-CT-CT-HC), and H-C-C-H (HC-CT-CT-HC) torsions from the
OPLS format given in Jorgensen et al, JACS 118, 11225 (1996) (which
are identical to those in Kony et al. J. Comput. Chem. 23 1416 (2002))
to the Ryckaert-Bellemans format given in ffoplsaabon.itp and found
that the calculated values are significantly different, particularly
for the C-C-C-C torsion.

A previous post to the gmx-users mailing list on March 27, 2008,
pointed out this issue for the H-C-C-H torsional potential but I did
not find any response to this post.

Does anyone know if there is an error in the ffoplsaabon.itp file? Or
is there a newer set of OPLS-AA parameters?

For the OPLS-AA parameters (in kcal/mol), I used:

dihedral              V1          V2             V3
C-C-C-C            1.740      -0.157         0.279
C-C-C-H             0.0          0.0            0.366
H-C-C-H             0.0          0.0            0.318

from which I calculated the Ryckaert-Bellemans parameters (in kJ/mol) as:

dihedral              C0          C1             C2           C3
C-C-C-C            3.56686   -1.889076   1.313776  -2.33467
C-C-C-H             0.66526   1.99577     0.0           -2.661024
H-C-C-H             0.76567   -2.29702    0.0           -3.06269

the parameters in the ffoplsaabon.itp file are:

dihedral              C0          C1             C2           C3
C-C-C-C           2.9288      -1.4644      0.2092     -1.6736
C-C-C-H           0.6276       1.88280    0.0           -2.5104
H-C-C-H           0.6276       1.8828      0.0           -2.5104

-David Huang