A Calculated Structures Database

[d3e102 "-at-" emsl.pnl.gov]

 A Computational Results Database
 Earlier today (August 12), Gijs Schaftenaar from the Dutch CAOS/CAMM Center
 asked whether there was any support for a "Calculated Structures
 Database"
 among the subscribers to the CCL. Here at the Environmental Molecular
 Sciences Laboratory we have had an on-going database project focused currently
 on ab initio electronic structure results.  We are actively working on
 making a first crude web-based interface available that would allow scientists
 to access optimized geometries, total energies, vibrational frequencies and
 relative energetics (e.g. dissociation energies or electron affinities) for
 208 molecules and the first 4 periods of the periodic table. A literature
 citation is associated with every result and the overwhelming majority of
 results have at least one experimental value available for comparision.
 Most of the popular Pople-style basis sets, plus the Huzinaga MINI and MIDI
 and correlation consistent basis sets are represented.  In terms of levels
 of theory, we have Hartree-Fock, MP2, MP3, MP4(SDTQ), CCSD, CCSD(T) and
 various kinds of CI.
 We have also developed a form that people could fill-in if they wish to submit
 the results of thier own calculations.  But a "gatekeeper" would have
 final
 say over whether externally contributed data made it into the database.
 I've spoken about and written about this project several timess and feel that I
 probably owe an explanation for the delay in getting it out.  It basically
 boils down to a matter of resources.  We are in the proccess of loading total
 energies and it is imperative that the numbers be correct.  That involves
 a lot of manual checking, which translates into human time. Simply stated,
 we won't make it available until we're convinced that the data is as
 correct as it's humanly possible to make it.
 After the release of the first version, which will have a limited number of
 ways that a user can interrogate the database, we plan to make available a
 second release with greatly enhanced flexibility and more types of
 information.
 This isn't meant to preclude any initiatives by the Dutch CAOS/CAMM Center.
 The task of building a database of computational results is a very
 daunting one.  There's lots for various groups to do.
 When the EMSL Computational Results Database becomes available, I will post
 a message to the CCL.
 --
 David Feller                                | Mail Stop K1-96
 Environmental Molecular Sciences Laboratory | 906 Battelle Blvd
 Battelle Pacific Northwest National Lab     | Richland, WA 99352
                                             |
 e-mail:d3e102 "-at-" emsl.pnl.gov                  | Fax: (509)-375-6631