THE MMCIF DICTIONARY: COMMUNITY REVIEW AND FINAL APPROVAL. Paula M. D. Fitzgerald, Merck Research Laboratories, Helen Berman, Department of Chemistry, Rutgers University, Philip Bourne, San Diego Supercomputing Center, Brian McMahon, International Union of Crystallography, Keith Watenpaugh, Physical and Analytical Chemistry, Pharmacia & Upjohn, and John Westbrook, Department of Chemistry, Rutgers University

The Crystallographic Information File (CIF) was developed by the IUCr Working Party on Crystallographic Information, in an effort sponsored by the IUCr Commission on Crystallographic Data and the IUCr Commission on Journals. The result of this effort, a dictionary of data items sufficient for archiving the small molecule crystallographic experiment and its results, was formally adopted by the IUCr in 1990.

In 1990, the IUCr formed a working group to expand the dictionary to include data items relevant to the macromolecular crystallographic experiment. As this effort progressed, we realized that the complex nature of the macromolecular experiment demanded a more rigorous data model than was provided for by the original CIF dictionary and its syntax laws (the Dictionary Description Language, DDL), and so a new DDL was developed and the mmCIF data model was recast as a flat-file representation of a relational database schema. This data model provides for the storage of information concerning all aspects of the macromolecular crystallographic structure determination process, beginning with the source of the material, and proceeding through crystallization, data collection, phasing, model fitting, model refinement and analysis, and description of the structure.

After five years of work and development, the macromolecular extensions to the CIF dictionary (mmCIF) were completed and presented to the community for review in August of 1995. The review process has resulted in a large number of changes, corrections, and additions, and we are extremely grateful to the many dedicated people who have looked carefully at the data model and its representation in the mmCIF dictionary, and who have made such cogent and thoughtful suggestions.

In March, 1996, the dictionary was opened to a wider audience for review and comment via announcements on several major bulletin boards. The dictionary itself, related documentation and examples, and related software and DDL information are publicly available on the World Wide Web at http://ndbserver.rutgers.edu/mmcif. Formal adoption of the dictionary by the IUCr is expected in mid-1996.