Help Using the PDB

Scope of the PDB Help Service

The PDB is a repository for macromolecular atomic structures and provides query capability for information on them - we do not have answers to questions on medical matters, the biology of particular proteins and cannot offer detailed software advice. You may try consulting our lists of other Web resources, education page and structural biology software lists to find information on some of these topics.

The PDB help service is provided by scientists and staff associated with the RCSB PDB. All questions receive a response and most are answered within 1-2 working days. Occasionally, when detailed expert advice needs to be obtained or PDB personel are away, the response time may be a little longer.

Frequently Asked Questions

FTP Services

Molecular Graphics

General Information

How to Download Structure Files A simple guide to downloading structure files from the Web is available here. This document provides a downloading procedure for individual files or a set of files that result from a specific query. If you need to obtain and store very large numbers of files (several hundreds) the Web download may be prohibitively large and you should consider obtaining the PDB holdings on CD-ROM. You may also download structure files from the PDB FTP site archive. There is no standard method for downloading a large number of structures with arbitrary PDB ids ( i.e. a set of structures that cannot be be obtained via a query). This task may be accomplished using a script to automate access and downloading from the FTP site - see here for an example.

Obtaining Complete Biological Assemblies Starting from the Structure Explorer page, click on the the Other Sources link. The PDB ID in the 'Quaternary Structure file' listing in the Structure Data section of the page links to the corresponding PDB entry in the EBI's Probable Quaternary Structure resource. PDB files that contain atomic coordinates expanded to form the complete likely quaternary structure may be downloaded from these pages.

Finding Structures in the PDB A simple introduction to finding proteins in the PDB is available. To find entries for a specific type of protein (for example, 'insulin') enter the protein name in the 'Compound Information' field in the SearchFields interface and then click on the 'Search' button. If a structure is not located it is worth trying a less specific search (for example, 'insulin' rather than 'porcine insulin') to avoid missing an entry when the search keywords are more specific than the PDB compound annotation. The less specific search might also find a structure that is similer enough (homologous) to be useful.

Searching for Ligands If you know the three-character heterogen (ligand) name then you can use the het: attribute in the SearchLite query engine or in the Text Search field in the SearchFields interface. For example, if the heterogen id is ABC you would enter het:ABC to find all PDB entries containing ABC coordinates. The current PDB het dictionary is obtainable in UNIX compressed CIF format from the PDB FTP archive at ftp://ftp.rcsb.org/pub/pdb/data/monomers/components.cif.Z.

It is also possible to use a text search using the common name for the ligand. This procedure is somewhat less reliable since it will pick up any instance of the ligand name. For example, you could search for 'benzamidine' and this would find all structures where this ligand was mentioned in the PDB entry text. However, this could miss some entries if the ligand was not mentioned by that name and could include some entries where benzamidine is a crystallization component but is not seen bound to the protein.

Searches and Lists of 'Representative Structures' The PDB query tools do not yet support queries limited to 'representative structure' subsets of the complete holdings. However, the OCA browser provides some capability for searches involving the representative structure subset. Lists of representative structures (by PDB ids) are available at the PDB Select and Culled PDB resources.

Uncompression of Files in the FTP Archive Files in the RCSB PDB FTP archive are in UNIX compressed form. Information on uncompression software is available. If you want to obtain uncompressed entries directly from the FTP server, omit the '.Z' portion from the file name.

Obtaining Files that used to be in the Brookhaven PDB FTP Archive The RCSB PDB is currently maintaining an up-to-date copy of the FTP archive with the same directory structure as the archive that used to be maintained by BNL. This archive is available at ftp://bnlarchive.rcsb.org.

Mirroring the Structure Holdings (PDB FTP archive) We welcome responsible mirroring of the PDB FTP site (i.e the archive containing the complete set of structure files in the PDB) for local use. Although we cannot offer individual assistance in establishing this type of mirror, policy and technical aspects of establishing this type of mirror are documented. It should be noted that no query software is provided with these mirrors. Users may wish to install the OCA browser (see here for an example) to provide a Web-based query service. OCA is designed to work in conjunction with the RCSB PDB FTP archive.

Problems with Rasmol Installation and Use Rasmol needs to be installed on your computer in order to use it via the Structure Explorer pages. Problems with the installation of Rasmol as a browser plug-in are very difficult to diagnose from the information that can usually be given by email. Common problems relate to outdated or unusual Web browsers, old versions of Rasmol and MIME types incorrectly set - some installation help is available. A troubleshooting guide is also available. Please note that although Rasmol is utilized by the PDB it is not PDB software and users are referred to the official Rasmol Web site for the most recent version and for more complete information. Another useful source of information on Rasmol and Chime applications is the Rasmol and Chime site maintained by Eric Martz.

Obtaining and Using Molecular Images from the PDB Users are free to use molecular images from the PDB site provided that a proper citation is a given. The VRML images that may be generated via the VRML (default options) links on the Structure Explorer/View Structure pages will not appear if you try to print the page directly from your browser. The simplest solution is to take a screen shot of the molecule and then print from the file containing this captured image. The VRML images generated via the custom interface do appear in a browser as VRML (.wrl) files which may be saved.

Obtaining Software for Molecular Graphics The PDB does not distribute software for molecular graphics but we maintain a list of widely used molecular graphics programs. These programs take PDB coordinate files as input. Instructions on downloading PDB files are available. Besides standalone use, in conjunction with a PDB coordinate files, the RasMol and Chime programs may be launched from a correctly configured browser via options on the PDB 'Structure Explorer/View Structure' pages.

Growth of the PDB Information on the growth in the number of structures available from the PDB is available.

Obtaining Documentation on the PDB File Format The latest version of the PDB file format is available for browsing (frames version) and may be downloaded as a single HTML document or as a plain text document. A PDB File Format FAQ is also available.

Obtaining Previous 'Molecules of the Month' Features Current and previous 'molecule of the month' features will continue to be available here.

Obtaining Sequence Data for the Entire PDB The pdb_seqres.txt file in the RCSB PDB FTP archive contains sequence data in FASTA format for all entries in the PDB. Each entry in this file corresponds is labeled with a PDB id and chain identifer. The sequence records in this file correspond to the SEQRES records in the PDB coordinate files. i.e. they are the complete primary sequences rather than the portions of the structure determined by x-ray crystallography or NMR spectroscopy.

Summaries of Structure Information Over the Whole PDB The PDB continues to maintain files listing all PDB ids versus author, compund and resolution, compound, crystal parameters, summary information, resolution and source information in a set of index files.