[Molecularmechanics] Molecular structure- Atoms and Sites

[Peterson,Brian K.]

I have a comment that mixes somewhat the discussions on force-field representation and structure representation.  Looking forward to new kinds of force-fields etc., might it be useful to describe structures as collections of sites and bonds rather than atoms and bonds?  Of course one can think of dummy-atom = site, but it seems conceptually more clear to me and potentially more flexible to have an atom being composed of at least one site and a molecule as a collection of atoms, sites, and bonds.  The most obvious use for such a distinction is to have sites which hold charge and which can be placed at essentially arbitrary locations.  An oxygen atom could consist of a nuclear site and sites for lone pairs etc.  These non-atom sites could also be useful since they will not have names in the periodic table and could easily be suppressed or not as desired in a gui.  They could also function as markers within a structure than might have nothing to do with the physics or chemistry, but would live within our structure representation.  Of course this adds complexity, but the flexibility might be worth it ????


brian


Brian K. Peterson, Ph.D.	Air Products and Chemicals, Inc.
<mailto:petersbk@apci.com>	7201 Hamilton Boulevard
phone: 610-481-3850		Allentown, PA  18195-1501
fax: 610-706-7672			USA


+