Jan K. Labanowski: Computational Portals for Chemistry  

Code Information Server

Many extensive software packages can run many different approaches, often differing in the number (and type/format) of files and parameters needed. For example, a well known chemistry code NWChem can perform small molecule quantum calculations on one hand, and the molecular dynamics using empirical potentials for macromolecules on the other. These computations have little in common. Which calculation is run depends on the input deck. From the functional point of view, these two modes of this software package, are two different programs with different requirements, i.e., correspond to two different AADs.

Informational resource which could translate science requirements to actual sofware package names, options, and files could be created in many ways. Vendors and experienced users can provide necessary maps. Moreover, convergence can be aided through various electronic communities and discussion forums, abd by forming standardizing consortia. But what needs to happen is generation of XML schemas which could describe software capabilities, and the creation of controlled dictionaries. This may unearth some old terminology wars (all quantum chemists know the MP2 versus MBPT debate, for example), but synonyms may be used.