Protein Data Bank (PDB)
RCSB PDB, a member of the Worldwide Protein Data Bank (wwPDB)
Every experimentally solved 3D protein structure since 1971, about 263,000 of them, free to anyone.
The PDB is the single global archive of experimentally determined 3D structures of proteins, nucleic acids, and complexes, solved by X-ray crystallography, cryo-EM, and NMR. Each entry gives atomic coordinates plus experimental details and resolution. It trained AlphaFold and underpins structural biology and drug design; RCSB adds search, visualization, and links to sequence and ligand data.
Access
- Cost
- FreeNo account; weekly releases, search and data APIs.
- License
- Public domainCC0 1.0 (wwPDB data)
- Sign-up
- No sign-up
- How you get it
- API and Bulk downloadPDBx/mmCIF, PDB, BinaryCIF
What's in it
- One row is
- One entry per deposited structure (four-character PDB ID)
- Covers
- Not geographic; depositors worldwide
- Years
- 1971-present
- Updated
- Weekly
- Size
- About 263,000 entries (November 2026 search API count)
Quality
Validated deposits with standard reports, maintained by a four-continent consortium. Coverage reflects what is easy to crystallize, not what is biologically important.
- Biased toward soluble, crystallizable proteins
- Older entries have lower resolution and validation standards
Connects to
- AlphaFold Protein Structure Database: Compare predicted models with experimental structures
- UniProtKB (UniProt Knowledgebase): Map structures to protein sequences and functions
Questions it could answer
- How fast has cryo-EM overtaken X-ray crystallography in new deposits?
- Which protein families remain structurally unsolved?