Home > Protein Data Bank

1 Contents

As of 1 October, 2004, the database contained 27,428 released atomic coordinate entries (or "structures") and took in about 2,000-3,000 new ones per year. Data is stored in the mmCIF format specifically developed for the purpose. Note that the database stores information about the exact location of all atoms in a large biomolecule; if one is only interested in sequence data, i.e. the list of amino acids making up a particular protein or the list of nucleotides making up a particular nucleic acid, the much larger databases from Swiss-Prot and the International Nucleotide Sequence Database Collaboration should be used.

Through the years the PDB has undergone many, many changes and revisions. Its original format was dictated by the width of computer punch cards.

• RCSB PDB Guide The PDB format specification can be found here, and it is vital that you read this before looking at the raw data.
• ftp.rcsb.org The raw data can be downloaded from here.
• www.rcsb.org Statistics about the PDB can be found here.

This legacy format has caused many problems with the format, and consequently the PDB has three distinct 'clean-up' projects;

• The Molecular Modeling DataBase (MMDB) from NCBI
• The Macromolecular Structure Database from the European Bioinformatics Institute
• The Data Uniformity Project from PDB

Each of these grant-funded projects has attempted to achieve the same goal via different routes. The Data Uniformity Project is hosted by the RCSB (the current home of the PDB). Each uses the original PDB data to derive a new format; The MMDB uses ASN.1 (and an XML conversion of this format); The MSD uses a Relational Database; The Data Uniformity Project uses mmCIF (and another XML conversion of this format).

Some people would say that this is a Good Thing; others would argue that, without a universal repository of information (i.e., a common dictionary), how can we talk about the same thing.

Each structure published in PDB receives a four-character alphanumeric identifier, its PDB ID. This should not be used as an identifier for biomolecules, since often several structures for the same molecule (in different environments or conformations) are contained in PDB with different PDB IDs.

If a biologist submits structure data for a protein or nucleic acid, PDB staff reviews and annotates it. The data is then automatically checked for plausibility. The source codeSource code (commonly just source or code is any series of statements written in some human-readable computer programming language. In modern programming languages, the source code which constitutes a software program is usually in several text files, but for this validation software has been released for free. The main data base accepts only experimentally derived structures, and not theoretically predicted ones (see protein structure predictionProtein structure prediction is one of the most significant tasks tackled in computational structural biology. It has the aim of determining the three-dimensional structure of proteins from their amino acid sequences. In more formal terms, this is the pre).

Various funding agencies and scientific journals now require scientists to submit their structure data to PDB.