What is the Protein Data Bank (PDB)?
Learning objectives
After reading this article you will be able to:
- Describe what the Protein Data Bank contains
- Compare the legacy PDB format with mmCIF
- Check a structure's quality before using it
The Protein Data Bank (PDB), founded in 1971, is the single worldwide archive of experimentally determined 3D structures of biological macromolecules. It is managed by the Worldwide PDB (wwPDB) partnership and holds more than 200,000 structures, each with a four-character identifier such as 1HSG.
What an entry contains
- Atomic coordinates for the protein, nucleic acids, bound ligands, ions and water.
- The experimental method (X-ray, cryo-EM, NMR) and quality statistics.
- Sequence, source organism, and literature references.
File formats: PDB and mmCIF
| Legacy PDB format (.pdb) | PDBx/mmCIF (.cif) | |
|---|---|---|
| Layout | Fixed-width text columns | Key–value and table-based text |
| Size limits | At most 99,999 atoms and 62 chains per file | No practical limit |
| Status | Still widely supported by tools | The wwPDB’s standard archive format |
Most docking programs still accept PDB files, while large complexes and modern structure-prediction outputs (such as those from AlphaFold3 and OpenFold3) are often delivered as mmCIF. Good viewers read both.
Choosing a structure for drug design
- Resolution. For X-ray structures, lower numbers are better; around 2.5 Å or better is usually suitable for docking.
- Ligand state. A structure with a ligand already bound in your site (a holo structure) usually docks better than an empty (apo) one.
- Completeness. Check for missing loops or side chains near the binding site.
- Relevance. Prefer the right species, mutation state and conformation for your project.
Last updated October 31, 2026
Put it into practice
Screen, dock and order compounds in one place
Pyxis Discovery combines macrocycle and small-molecule libraries with similarity search, docking and structure prediction in the browser.