The Protein Data Bank [PDB; Berman, Westbrook et al. (2000), Nucleic Acids Res. 28, 235-242; http://www.pdb.org/] is the single worldwide archive of primary structural data of biological macromolecules. Many secondary sources of information are derived from PDB data. It is the starting point for studies in structural bioinformatics. This article describes the goals of the PDB, the systems in place for data deposition and access, how to obtain further information and plans for the future development of the resource. The reader should come away with an understanding of the scope of the PDB and what is provided by the resource.

The Protein Data Bank (PDB; http://www.pdb.org/) is the international repository for three-dimensional structural data. The Nucleic Acid Database (NDB; http://ndbserver.rutgers.edu) specializes in nucleic acid-containing structures. These resources have developed several integrated tools for the deposition and query of structures in these databases.ADIT (AutoDep Input Tool) is used by the community for data validation and deposition, and internally by the PDB and NDB for processing and annotation. Because it is based on the data dictionary technology of mmCIF, ADIT is easily extended to accept and process information describing new science and technology. Examples of how this system is used for processing some of the most complex structures in the PDB and NDB will be described.The ADIT system has been designed to operate as a distributed network application or as a stand-alone tool. A workstation version of ADIT is available at http://deposit.pdb.org/software/ for researchers to prepare and validate entries in their home laboratories. This system also includes a variety of utility programs to assist in the extraction of information from several crystallographic applications, and utilities for merging the various program outputs into a single mmCIF data file ready for PDB deposition.The PDB is managed by three members of the RCSB: Rutgers, NIST, and SDSC. The PDB project is funded by the NSF, DOE, and two units of the NIH: NIGMS and NLM. The NDB is funded by the NSF and the DOE. The Protein Data Bank (PDB; http://www.pdb.org/) is the international repository for the processing and distribution of three-dimensional structural data. Its mission is to enable science by providing the most accurate and timely data for macromolecular structures. Data distribution and query functionality are replicated at six additional mirror sites, each of which maintains a Web site and an ftp archive. Weekly data updates are first tested on a local staging site, and then distributed to all production sites. All new functionality is first released on a public β test site (http://beta.rcsb.org/pdb/) prior to its distribution to all production sites. Examples of added query or display functionality include an enzyme classification browser, customized tabular reports, the pre-release of sequence information for some unreleased structures, and the STING Millennium Suite of graphical structure/sequence viewing tools (courtesy Goran Neshich and Barry Honig). Since the PDB holdings contain a considerable amount of redundancy, a sequence homology filter was implemented that provides the choice of displaying either a representative set of structures or the full search results. Progress on a re-engineering effort of the database, software, and Web interface will also be described. The PDB is managed by three members of the RCSB: Rutgers, NIST, and SDSC. The PDB project is funded by the NSF, DOE, and two units of the NIH: NIGMS and NLM. Keywords: DATABASES, STRUCTURAL BIOINFORMATICS, MMCIF Keywords: DATABASES, DATA DISTRIBUTION, QUERY FUNCTIONAL...