[gmx-users] Still force field issues: dihedral angle parameters

[Marc Baaden]

Hi,

I am still struggling with some force field parameter details.
Now it's the dihedrals. Take eg ffgmxbon.itp, you have:

[ dihedraltypes ]       ; func=2 below means improper dih potential 
  ; i    l func        q0          cq
 NR5*  NR5    2     0.000     167.360
...
[ dihedraltypes ]       ; func=1 below means GROMOS proper dih potential
  ; j    k func      phi0          cp   mult
    C   OA    1   180.000      16.736      2
..
[ dihedraltypes ]       ; func=3 below means RB potential
  CP2  CP2    3   9.2789   12.156  -13.120 -3.0597 26.240 -31.495


Given (as defined in the manual) a(n improper) dihedral angle potential
V(phi_ijkl),
why are i and l (2 external atoms ?) given in the first occurence of
"dihedraltypes" (impropers), j and k (2 internal atoms) in the second
occurence (proper periodic) and which ones for RB ?

This does not make sense for me. I would always expect 4 atom-type definitions
for i,j,k and l ..
.. unless you use general torsion parameters as in AMBER in which case
the 2 atoms should always be the central j/k ones as far as I understand
it. But then IMHO that would be non-sense with RB ..
What am I getting wrong ?

I noted that eg the "dihedrals" section has the full i,j,k,l atom type
definitions.
So what is the philosophy behind the general forcefield vs individual
molecule separation and the difference in the terms ?
(Seems not to be in the manual ..

Also, is it right that with proper periodic (func=1) potentials
you can only use one term of the cosine series expansion ?
So RB would be the only possible multiple-term implementation ?

Why does RB imply exclusion of 1,4 LJ interactions ? I cannot see any
theoretical reason (so is it for implementation purposes ?) ?


I think it would really help if all this was described in the manual,
and as a general rule for ALL force-field related things it would really
be utterly useful to have the units for each parameter given explicitly,
to avoid any confusion and minimize possible errors.
(don't get me wrong, I think the manual is already quite good and
comprehensive, but there is always this little bit that could be
improved .. and a clear definition of the forcefield is crucial if
you have to add/modify parameters from other sources)


Thanks in advance,
  Marc Baaden

-- 
 Dr. Marc Baaden - Laboratory of Molecular Biophysics, Oxford University
 mailto:baaden at smplinux.de  - ICQ# 11466242 -  http://www.marc-baaden.de
 FAX/Voice +49 697912 39550  -  Tel: +44 1865 275380  or  +33 609 843217