On Wed, Aug 11, 2010 at 7:26 AM, Chris Morley <c.mor...@gaseq.co.uk> wrote: > Although it has been touched on previously in the mailing lists, I > think we need to clarify what OBMol::_energy contains. It is accessed > in OBMol::SetEnergy(double) and by OBMol::GetEnergy(). > > Note that OBMol::GetEnergy(int) is separate, and is concerned with > relative of conformer energies. Its data is kept in OBConformerData. > > Can we agree that OBMol::_energy is the QM zero-point energy in > kcal/mole? (I take this to be the energy released on assembling the > molecule from elementary particles, but I may be wrong on this. > However it is a large negative number.) > > This arose because in gaussformat it was also being used for something > else (now commented out with Dave Lonie's agreement pending > resolution). It is also accessed in: > adfformat.cpp, cccformat.cpp, gamessformat.cpp, gulpformat.cpp, > mopacformat.cpp, mol2format.cpp, pwscfformat.cpp, vaspformat.cpp, > and (reading only) mmodformat.cpp, xyzformat.cpp, cmlformat.cpp > > Apart from the last (which I wrote), can those closer to these formats > confirm that the proposal above would be consistent with them? > > The comment in the declaration (and the API documentation) for > OBMol::GetEnergy()) is > "return the heat of formation for this molecule (in kcal/mol)" > I understand something different by heat of formation (thermodynamics, > defined to be zero for elements). Is "QM zero-point energy" the right > expression?
Another point that came up is what to do with other thermodynamic quantities. I use enthalpies frequently, as well as semiempirical/MM "energies" that have little to no physical meaning. I've been using a generic data pair to store these, but I think we should agree on some standard method for this, either a formal data structure (e.g. OBMol::_enthalpy, OBMol::_pm3, etc) or at least a set of "official" generic pair identifiers. With 2.3 pending, this might be a good time to go through and standardize the format readers. Dave ------------------------------------------------------------------------------ This SF.net email is sponsored by Make an app they can't live without Enter the BlackBerry Developer Challenge http://p.sf.net/sfu/RIM-dev2dev _______________________________________________ OpenBabel-Devel mailing list OpenBabel-Devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/openbabel-devel
