[gmx-users] Sugar atom types

Lieven Buts lieven at ultr.vub.ac.be
Fri Jun 14 13:58:16 CEST 2002

  In follow-up to my question about sugar topologies 
(2002/01/31) I have done some research into the parametrization
of GROMOS force fields for carbohydrates. Checking the literature
has left me with a number of questions.

Ott & Meyer (1995) [1] describe modifications of the GROMOS(87) 
force field to take into account the exo-anomeric [2] and gauche 
effects that occur in carbohydrates.

Kouwijzer et al. (1995) [3] describe an extension of GROMOS(87) to
an all-atom force field suitable for carbohydrates.

Finally, Spieser et al. (1999) [4] describe modifications to the 
GROMOS87 force field to improve the prediction of the energy difference
between different ring conformations and the conformer distribution
of the hydroxymethyl group. They also say the GROMOS96 force field
hardly differs from GROMOS87 as far as carbohydrates are concerned
(only the Van der Waals parameters of atom types CS1 and CS2 are
different), and that the same modifications could be applied to it.

I also notice that in ffgmx.rtp, glucose is defined using
CS1 and CS2 atoms for the carbon atoms, whereas in ffg43a1 force field,
it is defined using CH1 and CH2 atom types.

Does anyone know the exact origin of the gmx and G43a1 parameters
for carbohydrates, the differences between them and the relevance
of the modifications in the above-mentioned articles to them?


[1] Ott & Meyer (1996) 
    Parametrization of GROMOS force filed for oligosaccharides
    and assessment of efficiency of molecular dynamics simulations
    J. Comp. Chem. 17(8), 1068-1084.

[2]  The exo-anomeric effect is the result of a hyperconjugative
interaction between the exocyclic anomeric oxygen O1 and the
endocyclic sigma bond between O5 and C1:

      C5---O5    endocyclic
     / \    \ <- sigma bond
    /   H5   \     -
O4-C4-H4      C1---O1---R
|   \  H3 H2  /    -
HO4  \ /   \ /     ^
      C3---C2      |
      |     |      exocyclic oxygen atom with two electron pairs 
      O3    O2
      |     |
      HO3   HO2

  The results of this effect are shortening of the C1-O1 bond, 
lenghtening of the O5-C1 bond and a significant stabilisation 
of the gauche conformation with respect to the anti conformation.

[3] Kouwijzer,Van Eijck,Kooijman & Kroon (1995) 
    An extension og the GROMOS force field for carbohydrates,
    resulting in improvement of the crystal structure determination
    of alpha-D-galactose
    Acta Cryst. B51, 209-220.

[4] Spieser, van Kuik, Kroon-Batenburg & Kroon (1999) Improved
    carbohydrate force field for GROMOS: ring and hydroxymethyl 
    group conformations and exo-anomeric effect. 
    Carbohydrate Research 322, 264-273.

Lieven Buts
Vrije Universiteit Brussel

More information about the gromacs.org_gmx-users mailing list