Re: [sdpd] Which side should rotate ?

[MartinUlrich.Schmidt...@de.clariant.com]

An answer on the question "which side should rotate?"

> In Espoir 3.5, the choice is made, in case of a rotatable bond,
> to rotate the side of the bond with the less atoms. Is it a good
> idea ? I don't know, not enough time for testing. Alternative
> could be to rotate both sides randomly. But, it is believe that
> the larger side, of which a big part could be common for each
> torsion angle, may have such a weight that its rotation or the
> rotation of the whole molecule will have similar results. So
> that the random events are equally distributed between
> global rotation and rotation of a randomly selected torsion angle.
>
> Other ideas ?

>From my point of view there are generally 2 practicable definitions:

1) Keep that side fixed, which contains the Atom #1
(This is the usual definition, when a Z-matrix is used to describe a
molecular geometry)
2) Keep that side fixed, that appears first in the definition of the
torsion angles, e.g.
atom1-atom2-atom3-atom4 or atom4-atom3-atom2-atom1
in the sense of:
fixed atom - atom on rotation axis - atom on rotation axis - variable atom

Martin Schmidt


- - - - - - - -
Martin U. Schmidt
Clariant GmbH
Pigment Technology Research
G834
D-65926 Frankfurt a.M.
Germany
Tel.: 0049-69-305-13426
Fax: 0049-69-331749
E-Mail: MartinUlrich.Schmidt...@Clariant.com