[gmx-users] Improper dihedrals on planar rings
Justin A. Lemkul
jalemkul at vt.edu
Thu Oct 15 15:05:08 CEST 2009
Pablo Englebienne wrote:
> Hi all,
> I have a question about the setting of improper dihedrals. I have a
> pyrimidinone I want to use with the G53a5 force field, but I can't get
> the ring to stay flat.
> First question is: do all atoms in the planar ring need to have 3
> subsituents? I.e., should an explicit atom of type HC be present on the
> CHs within the ring, or could they be united atoms (e.g., non-bonded
> type CR1)?
There should be explicit HC. See, for example, the .rtp entry for PHE.
> Second question: how many parameters are required to keep the ring
> planar? I initially though that each bond within the ring should have an
> improper torsion defined, with the bonds in the J-K atoms in the
> torsion. In addition, every center bound to 3 other atoms (e.g., C=O,
> N-H) needs an additional term, with the central atom as atom I. If all
> atoms have 3 substituents, this would mean 6 bonds + 6 centers = 12
> improper torsions. Is that the minimum number of improper torsions
> required to keep a 6-membered ring flat? What about a 6-membered ring
> represented with united atoms?
Refer to manual section 4.2.11 and Figure 4.8. You can again also refer to the
PHE .rtp entry for an example of implementation. There are indeed multiple
impropers to define for each planar center, corresponding to the types shown in
the left and center panels of Figure 4.8.
I suppose it would be possible to define impropers for a united-atom 6-membered
aromatic ring (Figure 4.8, left), but this representation would be inconsistent
with the force field you wish to use. I am unsure why CR1 exists in 53a5; it
looks like it is used in a couple of cofactors, but in all the aromatic amino
acids, nucleotides, and heme, explicit HC are included on the aromatic rings.
> Thanks for reading so far (and many thanks for replying!).
Justin A. Lemkul
ICTAS Doctoral Scholar
Department of Biochemistry
jalemkul[at]vt.edu | (540) 231-9080
More information about the gromacs.org_gmx-users