RCSB PDB MCP for AI. Find structure details for proteins by ID or criteria.
Gemini Works with every AI agent you already use
…and any MCP-compatible client
Connect to your AI in seconds.
RCSB PDB connects your AI agent directly to the Protein Data Bank (PDB), the world's primary source for 3D molecular structures.
You search and retrieve experimental data on proteins, DNA, and RNA—everything from specific protein-ligand complexes to general structural metadata. Use it to find high-resolution structures by name, method (X-ray, Cryo-EM, NMR), or bound molecule ID.
It cuts out the manual database browsing required for structural biology research.
What your AI can do
Get pdb structure
PDB IDs are 4-character codes. This tool retrieves all structural details for a known PDB ID.
Search pdb
Searches the Protein Data Bank across 220K+ structures, returning key metadata like method, resolution, and ligands.
You search the entire database for molecules matching specific properties like method, resolution, or bound ligand.
The tool retrieves complete metadata and details for any known 4-character PDB ID.
You gather data points (resolution, method) across multiple entries to compare molecular architecture.
The search results specifically list bound molecules and their associated structural context.
Ask an AI about this
Waiting for input…
RCSB PDB MCP Server: 2 Tools for Structural Biology
Search the Protein Data Bank and retrieve detailed structural records for proteins, DNA, and RNA using two specialized tools.
Make your AI actually useful.
Add this MCP to Claude, Cursor, or Windsurf and your AI stops guessing. It gets real tools to look things up, take action, and handle the stuff you keep doing by hand.
Start using RCSB PDB on VinkiusGet Pdb Structure
PDB IDs are 4-character codes. This tool retrieves all structural details for a known PDB ID.
Search Pdb
Searches the Protein Data Bank across 220K+ structures, returning key metadata like...
Security and governance baked right in.
Pick your AI client below to get set up. Just create a Vinkius account, subscribe, and you're instantly up and running. We handle the entire backend infrastructure, delivering out-of-the-box support for HTTPS Streamable, SSE, and OAuth2—zero messy routing required.
Claude AI
Open Claude Settings
Go to claude.ai, click your profile icon, then navigate to Customize → Connectors.
Add Custom Connector
Click the "+" button and select Add custom connector. Paste your Vinkius endpoint URL:
https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp
Replace [YOUR_TOKEN_HERE] with your token
from cloud.vinkius.com. For OAuth-protected servers, expand
Advanced settings to add credentials.
Start a conversation
Open a new chat. The RCSB PDB integration is available immediately — no restart needed.
Choose How to Get Started
Build a custom MCP for your own tools, or connect a ready-made integration from our catalog.
Build Your Own
Turn any API into an MCP. Import a spec, define Agent Skills, or deploy with MCPFusion.
- Import from OpenAPI, Swagger, or YAML specs
- Create Agent Skills with progressive disclosure
- Deploy to edge with MCPFusion framework
- Built in DLP, auth, and compliance on every call
- Real time usage dashboard and cost metering
- Publish to catalog or keep private
Make Your AI Do More
Start with RCSB PDB, then connect any of our 5,100+ other servers whenever your AI needs more. One click, no limits.
- Use this MCP plus 5,100+ others, all in one place
- Add new capabilities to your AI anytime you want
- Every connection is secured and compliant automatically
- Track usage and costs across all your servers
- Works with Claude, ChatGPT, Cursor, and more
- New servers added to the catalog every week
Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by RCSB PDB. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.
VINKIUS INFRASTRUCTURE
Cloud Hosted
Managed infra
V8 Isolated
Sandboxed per request
Zero-Trust Proxy
No stored credentials
DLP Enforced
Policy on every call
GDPR Compliant
EU data residency
Token Compression
~60% cost reduction
Works with Claude, ChatGPT, Cursor, and more
The Model Context Protocol standardizes how applications expose capabilities to LLMs. Instead of operating in isolation, your AI gains direct access to external platforms, live data, and real-world actions through secure, standardized connections.
This connection provides 2 powerful capabilities that interface natively with Claude, ChatGPT, Cursor, and other compatible AI platforms. No middleware. No custom integration required.
Getting molecular structure metadata used to take pages of clicking and cross-referencing.
Today, figuring out if a specific protein complex was resolved by Cryo-EM or X-ray diffraction means navigating multiple database tabs. You find the ID, then click through resolution metrics, authors, ligands bound—it’s a frustrating process of copy-pasting and manual comparison across different institutional sites.
With this server, you ask your agent to search for criteria (e.g., 'protein with high Cryo-EM resolution'). The agent runs `search_pdb` and hands you a single result set that immediately gives the PDB ID, the method, *and* the resolution value—no clicks required.
RCSB PDB MCP Server: Get full structural records using get_pdb_structure.
Before this tool, getting a complete record for an established ID (like 1HHO) required visiting the source site and scrolling through detailed sections on chains, cofactors, and subunits. It was slow, tedious manual data entry.
Now, you just provide the PDB ID to `get_pdb_structure`. You get back a complete, structured report containing all that deep-dive metadata—everything about the polymer chains, the cofactor type, and the authors—in one clean output.
What your AI can actually do with this
You're connecting your AI agent straight into RCSB PDB, the biggest deal for 3D molecular structures. This isn't some flimsy database; it's the global hub where researchers dump all their experimental data on proteins, DNA, and RNA. If you gotta look up how big a molecule is or what kinda complex structure it forms, this tool set gets you there without having to manually click through twenty layers of menus.
search_pdb lets you search the entire Protein Data Bank—we're talking over 220,000 entries here. You don't just get a list; you can identify molecular structures by criteria like the experimental method used (was it X-ray Diffraction? Cryo-EM?), or what resolution was achieved. When you run this search, the results spit out key metadata: the protein or nucleic acid name and function, the precise resolution number, the associated authors, and when they dropped the data.
It's perfect for comparing structures; you can pull data points—like resolution and method—across multiple entries to compare molecular architecture on the fly. The search results also specifically list any bound molecules or cofactors, so if you gotta know what ligand is attached, it shows up right there in the metadata.
If you already know exactly which structure you're looking at, get_pdb_structure is your shortcut. You just drop a 4-character PDB ID—like '1HHO' or whatever's on your list—and it pulls the complete structural record instantly. This tool retrieves all the granular details: everything from polymer chain counts to specific cofactor binding sites.
When you use these tools, your agent reads deep into the molecular data. You can find structures by name, method, or even by a bound molecule's ID. It bypasses all that tedious database browsing and gives you raw structural facts immediately.
019d75fc-c796-7011-9988-b4987a177ff6 
Here's how it actually works
The bottom line is: You query by criteria first, then retrieve the deep data point using the ID provided in the results.
Send a query specifying what you need (e.g., 'Find all Cryo-EM structures of ribosomes').
The agent uses search_pdb to narrow the search, returning relevant PDB IDs and summarizing their key characteristics.
If results are found, or if you provide an ID, the system executes get_pdb_structure to pull the full, deep-dive record for that specific entry.
Who is this actually for?
Structural biologists and computational chemists use this when they need to move beyond general literature reviews. They aren't guessing; they're pinpointing specific molecular interactions—like a drug binding site or a protein domain structure. This saves them hours of manual cross-referencing in database interfaces.
They use the server to quickly validate structural hypotheses, checking if similar proteins have been resolved using specific methods or resolutions.
They search for protein-ligand co-crystal structures (using search_pdb) needed for structure-based drug design and virtual screening efforts.
They use the tool to explore real, annotated experimental data, helping them understand complex protein folding and architecture in a controlled environment.
What Changes When You Connect
Saves time on metadata checks. Instead of visiting multiple database tabs, search_pdb gives you one summary record listing the method, resolution, and authors right away.
Directly access complex structures. Use get_pdb_structure with a 4-character PDB ID (like 6LU7) to pull full data on everything from cofactors to polymer chains.
Focuses drug design efforts. You find protein-ligand complexes specifically by searching, bypassing general searches and getting right to the structural details needed for virtual screening.
Handles diverse molecular types. The server supports search across proteins, DNA, and RNA structures, making it useful regardless of your macromolecule focus.
Bypasses API key issues. You start searching immediately without needing to manage or generate external API keys.
See it in action
Evaluating a novel drug target's structure
A chemist knows the protein name but not its ID. They ask their agent, which runs search_pdb to find candidates matching 'kinase inhibitor'. The results narrow down potential targets and provide crucial metadata like resolution (e.g., 2.5 Å) for assessing data quality.
Checking a specific structure's details
A student is given PDB ID '1HHO'. They use get_pdb_structure to instantly pull all structural information, verifying it’s deoxy human hemoglobin and seeing the method (X-ray Diffraction) and resolution.
Comparing multiple reaction mechanisms
A researcher needs structures from different sources. They use search_pdb to find all Cryo-EM entries related to 'ribosome complexes,' allowing them to compare the molecular machine's structure across various experimental conditions.
Validating binding site information
A biophysicist wants to know if a specific ligand (e.g., N3) was found in high-resolution studies. They use search_pdb and filter by 'bound ligands' to quickly identify relevant PDB entries.
The honest tradeoffs
Searching only by protein name
A user simply asks the agent, 'Tell me about hemoglobin.' This is too general and will return hundreds of ambiguous results without context.
Don't just provide a name. Use search_pdb and specify constraints: 'Search for deoxy human hemoglobin using X-ray Diffraction to get its PDB ID.'
Assuming the ID is correct
A user types an invalid or outdated 4-character code into a retrieval tool, wasting time on a non-existent data point.
Always run search_pdb first to validate and confirm the most current PDB IDs for your target molecule before attempting structural lookups.
Using basic search engines
Searching Google for 'SARS-CoV-2 structure' only yields links, requiring multiple clicks and manual data extraction.
Use search_pdb to query the database directly. The agent will pull structured results detailing method (X-ray), resolution (e.g., 2.16 Å), and ligands in one step.
When It Fits, When It Doesn't
Use this server if your primary need is molecular structure data from the PDB—specifically, you are comparing resolutions, identifying binding sites, or confirming structural metadata for proteins, DNA, or RNA. You must know that the core data comes from experimental methods like X-ray crystallography or Cryo-EM.
Don't use this if you just need general biological knowledge (e.g., 'What is glycolysis?'). For those questions, a general LLM works fine. Also, don't use it if your query requires data interpretation (e.g., 'Why did the binding change?'); this tool only provides raw structural data points. Use search_pdb when you need to narrow down a field of possibilities; save get_pdb_structure for when you have the precise 4-letter identifier.
Questions you might have
How do I find structures using the RCSB PDB MCP Server? +
Use search_pdb. You need to tell it criteria like 'protein name,' 'method,' or 'ligand' so it can search the 220K+ entries and give you relevant IDs.
What is the difference between using get_pdb_structure vs search_pdb? +
search_pdb finds candidates based on criteria. get_pdb_structure requires a specific, known 4-character PDB ID and pulls the full record for that single entry.
Can I find structures for DNA or RNA? +
Yes. The server supports searching across proteins, DNA, and RNA molecules; you just need to include the type in your query when using search_pdb.
Does RCSB PDB MCP Server require an API key? +
No. You can start querying molecular structures immediately without needing to manage or generate any API keys.
When I use `get_pdb_structure`, what metadata does it provide about experimental resolution? +
The tool provides the exact recorded resolution (e.g., 2.16 Å) alongside the structural details. This metric tells you the precision of the original physical measurement, which is crucial when comparing different structures.
If I use `search_pdb` with a very broad query, how can I optimize the search for speed? +
To keep things fast, always narrow your criteria by including restrictive filters first. Combining specific methods (like Cryo-EM) or known ligands dramatically reduces the dataset from 220K+ entries.
What happens when `get_pdb_structure` is called with a PDB ID that doesn't exist? +
The system returns an explicit error stating the PDB ID was not found. Your AI client can catch this specific failure and prompt you to re-check the 4-character code or adjust your search parameters.
When using `search_pdb`, how do I ensure I only find protein-ligand co-crystal structures? +
You must include 'bound ligand' or 'complex structure' in the tool's criteria. The resulting data will list the associated ligands and their binding site details, confirming a complex structure.
What experimental methods are represented in the PDB? +
The PDB contains structures determined by four main methods: X-ray Crystallography (85% of entries, atomic resolution), Cryo-EM (rapidly growing, now 15%+ of new deposits), NMR Spectroscopy (solution-state structures), and Neutron Diffraction (hydrogen atom positions). Each entry lists its method and resolution.
Do I need authentication to access PDB data? +
No. RCSB PDB is completely free and open — funded by NSF, NIH, DOE, and international partners. All 220K+ structures are openly accessible without any registration or API key. The data is licensed under CC0 (public domain).
What information does each PDB entry contain? +
Each entry includes: PDB ID (4-character code), structure title, depositing authors, experimental method, resolution in Ångströms, source organism, release date, number of polymer chains, and bound non-polymer ligands (drugs, cofactors, ions). Higher resolution (lower number) means more precise atomic coordinates.
We've already built the connector for RCSB PDB. Just plug in your AI agents and start using Vinkius.
No hosting. No infrastructure. No complex setup.
All 2 tools are live and waiting.
You're up and running in seconds.
Gemini Vinkius gives your AI agents access to the full catalog of app connectors, all fully managed, secure, and enterprise-ready. One subscription, every tool you need.
Built, hosted, and secured by Vinkius. You just connect and go.