The Protein Data Bank (PDB) is the single global archive for experimentally determined three-dimensional (3D) structures of proteins, nucleic acids, and complex assemblies. Established in 1971 at Brookhaven National Laboratory and now managed by the Worldwide Protein Data Bank (wwPDB) consortium, it stores atomic coordinates, experimental metadata, and validation reports derived primarily from X-ray crystallography, nuclear magnetic resonance (NMR) spectroscopy, and cryo-electron microscopy (cryo-EM).
Glossary
Protein Data Bank (PDB)
The Protein Data Bank (PDB) is the single worldwide repository for experimentally determined 3D structures of biological macromolecules, providing the foundational training data for protein structure prediction models.
GLOBAL BIOMOLECULAR ARCHIVE
What is the Protein Data Bank (PDB)?
The definitive, single worldwide repository for experimentally determined 3D structures of large biological molecules, serving as the foundational training data for modern protein structure prediction models.
As the primary source of ground-truth structural data, the PDB is indispensable for training and benchmarking deep learning models like AlphaFold and RoseTTAFold. The archive's rigorous curation and standardized .pdb and .mmCIF file formats provide the high-fidelity (x, y, z) coordinate data required for supervised learning, enabling the prediction of residue coevolution and the validation of computationally designed de novo proteins against empirical reality.
THE WORLDWIDE PROTEIN DATA BANK
Key Features of the PDB
The Protein Data Bank (PDB) is the single global repository for experimentally determined 3D structures of biological macromolecules. Established in 1971, it serves as the foundational training data for all modern protein structure prediction models, including AlphaFold.
01
Experimental Structure Repository
The PDB archives atomic coordinates derived from three primary experimental methods:
- X-ray Crystallography: The dominant method, providing high-resolution structures by analyzing diffraction patterns from protein crystals.
- Cryo-Electron Microscopy (Cryo-EM): Rapidly growing in contribution, capturing macromolecules in near-native, solution-like states without crystallization.
- Nuclear Magnetic Resonance (NMR) Spectroscopy: Provides dynamic information about protein structures and flexibility in solution.
Each entry is assigned a unique 4-character PDB ID (e.g., 1BNA, 7R6R) and includes primary citation metadata, sequence information, and atomic coordinate files.
215,000+
Total Structures
1971
Year Established
03
Global Validation Metrics
Every PDB entry undergoes rigorous validation and is accompanied by a slider-based graphical report assessing geometric quality. Key metrics include:
- Ramachandran Plot Analysis: Quantifies the percentage of residues in favored, allowed, and outlier regions of phi/psi dihedral angle space.
- Clashscore: The number of serious steric overlaps (van der Waals violations) per 1,000 atoms, with lower scores indicating better geometry.
- RSRZ (Real-Space R-value Z-score): Measures per-residue fit of the atomic model to the experimental electron density map.
- Rotamer Outliers: Identifies side-chain conformations that deviate significantly from statistically preferred rotamer libraries.
These metrics are critical for filtering high-quality training data for AlphaFold and RoseTTAFold pipelines.
>98%
Ramachandran Favored (Avg)
<5
Avg Clashscore
05
Biological Assembly and Quaternary Structure
The asymmetric unit deposited in the PDB is often not the functional biological molecule. The PDB provides generated biological assemblies by applying crystallographic and non-crystallographic symmetry operations.
- Stoichiometry: The precise subunit composition (e.g., homodimer, heterotetramer).
- Author-Defined Assembly: The oligomeric state specified by the depositing researchers.
- PISA (Proteins, Interfaces, Structures and Assemblies): An automated algorithm that predicts the most thermodynamically stable quaternary structure based on buried surface area and solvation energy.
Understanding biological assemblies is essential for training quaternary structure prediction models and analyzing protein-protein interaction interfaces.
~50%
Entries are Oligomeric
06
Ligand and Cofactor Annotation
The PDB contains extensive information on non-polymer chemical components bound to proteins:
- HETATM Records: Atomic coordinates for ligands, cofactors, ions, and solvent molecules.
- 3-Letter Residue Codes: Unique identifiers for each chemical component (e.g., ATP, HEM, NAG).
- Chemical Component Dictionary (CCD): A reference database defining ideal bond lengths, angles, and stereochemistry for every ligand.
- Binding Affinity Data: Often linked to external resources like PDBbind, which curates experimentally measured Kd, Ki, and IC50 values.
This ligand data is the primary source for training drug-target interaction prediction models and molecular docking scoring functions.
Enabling Efficiency, Speed & Accuracy
Intelligent Analysis, Decision & Execution
We build AI systems for teams that need search across company data, workflow automation across tools, or AI features inside products and internal software.
Talk to Us
Search across company data
Give teams answers from docs, tickets, runbooks, and product data with sources and permissions.
Useful when people spend too long searching or get different answers from different systems.
Enterprise searchRAGPermissions
Read more
Automate internal workflows
Use AI to route work, draft outputs, trigger actions, and keep approvals and logs in place.
Useful when repetitive work moves across multiple tools and teams.
AI agentsWorkflow automationGovernance
Read more
Add AI to products and internal tools
Build assistants, guided actions, or decision support into the software your team or customers already use.
Useful when AI needs to be part of the product, not a separate tool.
AI integrationDecision supportModel routing
Read morePROTEIN DATA BANK ESSENTIALS
Frequently Asked Questions
Clear answers to the most common questions about the Protein Data Bank, the foundational archive for structural biology and AI-driven protein structure prediction.
The Protein Data Bank (PDB) is the single, global archive for experimentally determined 3D structures of biological macromolecules, including proteins, nucleic acids, and complex assemblies. It works as a centralized repository where researchers deposit atomic coordinate data derived primarily from X-ray crystallography, nuclear magnetic resonance (NMR) spectroscopy, and cryo-electron microscopy (cryo-EM). Each entry is assigned a unique 4-character alphanumeric identifier (e.g., 1BNA, 7R6R) and contains primary atomic coordinates, metadata about the experimental conditions, and annotations on secondary structure and ligand binding. The archive, managed by the worldwide Protein Data Bank (wwPDB) consortium, serves as the ground-truth training dataset for deep learning models like AlphaFold and RoseTTAFold, making it the indispensable empirical foundation for all modern computational structural biology.
About the author
Prasad Kumkar
CEO & MD, Inference Systems
Prasad Kumkar is the CEO & MD of Inference Systems and writes about AI systems architecture, LLM infrastructure, model serving, evaluation, and production deployment. Over 5+ years, he has worked across computer vision models, L5 autonomous vehicle systems, and LLM research, with a focus on taking complex AI ideas into real-world engineering systems.
His work and writing cover AI systems, large language models, AI agents, multimodal systems, autonomous systems, inference optimization, RAG, evaluation, and production AI engineering.
LinkedIn
Limited slotsGet a Free AI Consultation
Partnered with leading AI, data, and software stack.
How We Work
Custom AI workflows for your Business
One-fit-all AI don't work for modern businesses. At Inferensys, we aim to understand your business & custom requirements; which we use to define most efficient agentic workflows, the data, and the tools for your business.
01
Review the use case
We understand the task, the users, and where AI can actually help.
Read more02
Pick the right approach
We define what needs search, automation, or product integration.
Read more03
Build the first useful version
We implement the part that proves the value first.
Read more04
Improve from there
We add the checks and visibility needed to keep it useful.
Read moreThe first call is a practical review of your use case and the right next step.
Talk to Us