Rosetta MCP Server
An MCP server that enables AI coding assistants to interact with Rosetta, PyRosetta, and Biotite for running RosettaScripts, validating XML, translating between Rosetta and Biotite, scoring structures, and querying documentation.
README
<p align="center"> <img src="./assets/banner2.png" width="100%" alt="Rosetta MCP Server" /> </p>
Rosetta MCP Server
Author: Ariel J. Ben-Sasson
A Model Context Protocol (MCP) server that lets Cursor (or any MCP client) work with Rosetta, PyRosetta, and Biotite: run RosettaScripts, validate XML protocols, translate between Rosetta and Biotite, score structures, and query documentation -- all from your AI coding assistant.
What's new in v1.3.0 (vs v1.1.8 on npm)
New: Biotite integration
rosetta_to_biotite-- Find the Biotite equivalent of any Rosetta function with working example code (21 mappings covering structure I/O, SASA, RMSD, superimposition, secondary structure, contacts, hydrogen bonds, B-factors, angles, and more)biotite_to_rosetta-- Reverse lookup: find the Rosetta equivalent of a Biotite functiontranslate_rosetta_script_to_biotite-- Translate entire RosettaScripts XML or PyRosetta code to Biotite Python. Design/optimization operations are flagged as Rosetta-only.- Fuzzy search with keyword aliases ("contacts", "binding energy", "surface area", "align", etc.)
Improved: XML to PyRosetta translator
- 37 element types supported (was 6): 11 movers, 9 filters, 10 selectors, 7 task operations
- Full attribute handling:
repeats,disable_design,cartesian,tolerance,threshold,distance, and more - Child element support:
MoveMap(withSpan),Reweight,ScoreFunction - Reports unrecognized elements so you know what needs manual work
Improved: Help and documentation
get_rosetta_helpnow accepts any topic: movers by name ("FastRelax"), concepts ("constraints", "docking"), or score functions ("ref2015") -- auto-fetches live docs from rosettacommons.orgsearch_rosetta_web_docsfallback: when DuckDuckGo is rate-limited, probes direct Rosetta docs URLsget_cached_docsauto-caches: no need to callcache_cli_docsfirst- Expanded static help for score_functions, movers, filters, xml, and parameters
Improved: Scoring
pyrosetta_score: newper_residueoption returns per-residue energy breakdownscorefxnparameter now works (was ignored in v1.1.8)- Proper error messages for missing files instead of silent
{}
Improved: Validation
validate_xml: newvalidate_against_schemaoption checks element names against the Rosetta XSD schema (catches typos likeFastRleax)
MCP spec compliance fixes
tools/callresponses now use correct{ content: [{ type: "text", text }] }format- Tool errors return
isError: true(not JSON-RPC errors) - Standard JSON-RPC error codes (-32601, -32700, -32603)
- Removed false
resourcescapability advertisement
Security fixes
- User input no longer interpolated into Python code (uses env vars / stdin)
- Temp files written to
os.tmpdir()(not module directory)
Cleanup
- Removed 3 redundant tools:
list_functions(merged intoget_rosetta_info),search_pyrosetta_wheels,cache_cli_docs(auto-cache inget_cached_docs) - Removed hardcoded personal paths
- Fixed shadowed variables, async anti-patterns, dead code
- 18 tools (was 21), all with improved agent-oriented descriptions
Example: asking a naive question
This is what makes the MCP server powerful -- an AI agent can answer domain questions by calling the right tools automatically:
User asks in Cursor: "How do I relax my protein and what's the Biotite equivalent?"
The agent calls two MCP tools behind the scenes:
1. get_rosetta_help("FastRelax") returns 6000+ chars of live documentation:
FastRelax performs all-atom relaxation using the FastRelax protocol. Parameters include
scorefxn,repeats,cartesian,disable_design,MoveMapconfiguration...
2. rosetta_to_biotite("FastRelax") returns:
{
"found": true,
"results": [{
"rosetta": { "name": "FastRelax", "example": ["relax = FastRelax()", "relax.set_scorefxn(get_score_function('ref2015'))", "relax.apply(pose)"] },
"biotite": null,
"equivalence": "none_from_biotite",
"notes": "Biotite does NOT perform structure optimization. These are Rosetta-specific capabilities."
}]
}
The agent synthesizes: "FastRelax is Rosetta's all-atom relaxation protocol. Here's how to use it... Note: Biotite is analysis-only and has no equivalent -- you need PyRosetta for structure optimization."
Without the MCP, the agent would guess from training data and likely get parameter names or API signatures wrong.
What you get (18 tools)
Discovery & Help
| Tool | Description |
|---|---|
get_rosetta_info |
All available score functions, movers, filters, selectors, parameters |
get_rosetta_help |
Help for any topic -- accepts mover names, concepts, or score functions |
pyrosetta_introspect |
Live PyRosetta API search with docs and signatures |
Documentation
| Tool | Description |
|---|---|
search_rosetta_web_docs |
Search rosettacommons.org documentation |
get_rosetta_web_doc |
Fetch and read a specific docs page |
get_cached_docs |
Search cached CLI help (auto-caches on first use) |
Execution & Scoring
| Tool | Description |
|---|---|
run_rosetta_scripts |
Run a RosettaScripts XML protocol on a PDB |
pyrosetta_score |
Score a PDB with optional per-residue breakdown |
Translation
| Tool | Description |
|---|---|
xml_to_pyrosetta |
XML to PyRosetta Python (37 element types) |
rosetta_to_biotite |
Find Biotite equivalent of a Rosetta function |
biotite_to_rosetta |
Find Rosetta equivalent of a Biotite function |
translate_rosetta_script_to_biotite |
Translate full scripts from Rosetta to Biotite |
Validation & Schema
| Tool | Description |
|---|---|
validate_xml |
Check XML syntax + optional schema validation |
rosetta_scripts_schema |
Generate XSD schema and extract element names |
Environment
| Tool | Description |
|---|---|
python_env_info |
Python version and installed packages |
check_pyrosetta |
Verify PyRosetta is available |
install_pyrosetta_installer |
Auto-install PyRosetta (10-30 min) |
find_rosetta_scripts |
Locate the rosetta_scripts binary |
Quick start
1. Install from npm
npm install -g rosetta-mcp-server
2. Set up Python environment
# Create a venv with PyRosetta and Biotite
uv venv ~/.venvs/rosetta-mcp
~/.venvs/rosetta-mcp/bin/pip install pyrosetta-installer biotite
~/.venvs/rosetta-mcp/bin/python -c "import pyrosetta_installer as I; I.install_pyrosetta()"
Or skip this step -- PyRosetta auto-installs on first use (takes 10-30 min).
3. Configure your MCP client
Cursor (~/.cursor/mcp.json):
{
"mcpServers": {
"rosetta": {
"command": "rosetta-mcp-server",
"args": [],
"env": {
"ROSETTA_BIN": "/path/to/rosetta_scripts.default.macosclangrelease",
"PYTHON_BIN": "/path/to/.venvs/rosetta-mcp/bin/python"
}
}
}
}
Claude Desktop (~/Library/Application Support/Claude/claude_desktop_config.json):
{
"mcpServers": {
"rosetta": {
"command": "rosetta-mcp-server",
"env": {
"ROSETTA_BIN": "/path/to/rosetta_scripts.default.macosclangrelease",
"PYTHON_BIN": "/path/to/.venvs/rosetta-mcp/bin/python"
}
}
}
}
Environment variables:
| Variable | Required | Description |
|---|---|---|
ROSETTA_BIN |
No | Path to rosetta_scripts binary or its directory. If not set, searches common paths and PATH. |
PYTHON_BIN |
No | Python interpreter with PyRosetta/Biotite. Defaults to python3. |
MCP_DEBUG |
No | Set to 1 for debug logging to stderr. |
4. Restart your editor
Open Settings -> MCP. The "rosetta" server should appear green with 18 tools.
XML to PyRosetta translation example
Input XML:
<ROSETTASCRIPTS>
<SCOREFXNS>
<ScoreFunction name="ref" weights="ref2015"/>
</SCOREFXNS>
<RESIDUE_SELECTORS>
<Chain name="chainA" chains="A"/>
</RESIDUE_SELECTORS>
<MOVERS>
<FastRelax name="relax" scorefxn="ref" repeats="5" cartesian="true"/>
</MOVERS>
<PROTOCOLS>
<Add mover="relax"/>
</PROTOCOLS>
</ROSETTASCRIPTS>
Generated PyRosetta code:
import pyrosetta
from pyrosetta import pose_from_pdb
from pyrosetta.rosetta.core.scoring import get_score_function
from pyrosetta.rosetta.core.select.residue_selector import *
from pyrosetta.rosetta.protocols.relax import *
pyrosetta.init("-mute all")
pose = pose_from_pdb("your_protein.pdb")
# Residue Selectors
chainSelector = ChainSelector()
chainSelector.set_chain_strings("A")
# Movers
fastRelax = FastRelax()
fastRelax.set_scorefxn(get_score_function("ref"))
fastRelax.set_default_repeats(5)
fastRelax.cartesian(True)
sfxn = get_score_function("ref2015")
# Apply movers
fastRelax.apply(pose)
pose.dump_pdb("output.pdb")
score = pose.energies().total_energy()
print(f"Final score: {score}")
Rosetta <-> Biotite mapping coverage
| Category | Rosetta | Biotite | Equivalence |
|---|---|---|---|
| Structure I/O | pose_from_pdb |
PDBFile.read |
Full |
| Structure I/O | pose.dump_pdb |
PDBFile.write |
Full |
| Structure I/O | pose_from_file (CIF) |
CIFFile.read |
Full |
| Surface Analysis | SasaMetric |
biotite.structure.sasa |
Full |
| Alignment | SuperimposeMover |
biotite.structure.superimpose |
Full |
| RMSD | all_atom_rmsd |
biotite.structure.rmsd |
Full |
| Secondary Structure | DsspMover |
annotate_sse |
Partial |
| Sequence | pose.sequence() |
get_residues |
Full |
| Distance | AtomPairConstraint |
biotite.structure.distance |
Full |
| Angles | pose.phi/psi/omega |
biotite.structure.dihedral |
Full |
| Interface | InterfaceAnalyzerMover |
sasa + selection |
Partial |
| Database | rcsb.pose_from_rcsb |
rcsb.fetch |
Full |
| Selection | ChainSelector etc. |
numpy boolean indexing | Full |
| Contacts | distance matrices | CellList |
Partial |
| Ramachandran | pose.phi/psi |
dihedral_backbone |
Partial |
| H-bonds | HBondSet |
biotite.structure.hbond |
Partial |
| B-factors | pdb_info().bfactor |
AtomArray.b_factor |
Full |
| Center of Mass | center_of_mass |
mass_center |
Full |
| Scoring | ScoreFunction |
None | Rosetta only |
| Optimization | FastRelax |
None | Rosetta only |
| Design | FastDesign |
None | Rosetta only |
Troubleshooting
- Server shows red in Cursor: Restart Cursor. Use absolute path in config (e.g.,
/opt/homebrew/bin/rosetta-mcp-server). Ensure Node 14+ and Python 3.8+. run_rosetta_scriptsfails: VerifyROSETTA_BINpoints to a valid binary. Try"$ROSETTA_BIN" -help.- PyRosetta tools say "not available": Install via
pip install pyrosetta-installerthen run the installer, or let the MCP server auto-install on first use. - Biotite tools return no results: Install Biotite in the same Python env:
pip install biotite get_rosetta_helpreturns "No detailed help": Try the exact Rosetta class name (e.g., "FastRelax" not "relax"). The tool resolves common aliases but may miss unusual names.
Verify from the command line
# Check version
echo '{"jsonrpc":"2.0","id":1,"method":"initialize","params":{"protocolVersion":"2024-11-05"}}' | rosetta-mcp-server 2>/dev/null | python3 -c "import sys,json; print(json.loads(sys.stdin.readline())['result']['serverInfo'])"
# List all tools
echo '{"jsonrpc":"2.0","id":1,"method":"tools/list","params":{}}' | rosetta-mcp-server 2>/dev/null | python3 -c "import sys,json; [print(t['name']) for t in json.loads(sys.stdin.readline())['result']['tools']]"
Development
rosetta-mcp-server/
├── rosetta_mcp_wrapper.js # Node MCP server (protocol + all 18 tools)
├── rosetta_mcp_server.py # Python helper (static Rosetta data)
├── install_pyrosetta.js # Standalone PyRosetta installer
├── package.json # npm package config
└── README.md
License and attribution
- MIT for this repository
- Rosetta/PyRosetta: see RosettaCommons licenses; commercial use requires the appropriate license
- Biotite: BSD 3-Clause license
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。