MCP server for making it easy to run Rowan's molecular design and simulation tools.
Just add this to your MCP configuration and it will automatically install and run:
HTTP/SSE configuration:
{
"mcpServers": {
"rowan": {
"type": "http",
"url": "http://127.0.0.1:6276/sse"
}
}
}Then start the server:
# Set your API key
export ROWAN_API_KEY="your_api_key_here"
# Start the HTTP server
uvx --from rowan-mcp rowan-mcpIf you prefer to install the package first:
Using uv:
uv add rowan-mcpUsing pip:
pip install rowan-mcpThen configure and start:
{
"mcpServers": {
"rowan": {
"type": "http",
"url": "http://127.0.0.1:6276/sse"
}
}
}# Set API key and start server
export ROWAN_API_KEY="your_api_key_here"
rowan-mcpVisit labs.rowansci.com → Create account → Generate API key
Ask your AI: "Calculate the pKa of aspirin" or "Optimize the geometry of caffeine"
Ask the LLM to:
- Calculate drug properties: "Predict drug-likeness of aspirin"
- Optimize molecular structures: "Optimize the geometry of aspirin"
- Predict chemical behavior: "What's the pKa of acetic acid?"
- Run calculations: "Calculate the HOMO and LUMO of benzene"
- Python 3.11+
- Package manager: uv (recommended) or pip
- Rowan API key (free at labs.rowansci.com)
- MCP-compatible client (Claude Desktop, etc.)
Development commands (if you cloned the repo):
# Run from source
export ROWAN_API_KEY="your_api_key_here"
uv run python -m rowan_mcpTotal: 45 MCP Tools
- 28 Dedicated Workflow Functions (all workflows from Rowan v2.1.9)
- 1 Batch Workflow Function
- 3 Molecule Lookup Tools
- 7 Workflow Management Tools
- 6 Protein Management Tools
submit_basic_calculation_workflow- Energy, optimization, frequencies with multiple engines (omol25, xtb, psi4)submit_conformer_search_workflow- Conformational search with multiple search modes (rapid/careful/meticulous)submit_conformers_workflow- Conformer generation and enumeration (different from conformer_search)submit_multistage_opt_workflow- Multi-stage geometry optimization with sequential accuracy levelssubmit_scan_workflow- Molecular scans (dihedral, bond, angle) with wavefront propagationsubmit_irc_workflow- Intrinsic reaction coordinate calculations for transition states
submit_pka_workflow- Microscopic pKa calculations with customizable pH ranges and elementssubmit_macropka_workflow- Macroscopic pKa calculations across pH and charge rangessubmit_solubility_workflow- Solubility predictions across multiple solvents and temperaturessubmit_redox_potential_workflow- Electrochemical reduction/oxidation potentialssubmit_descriptors_workflow- ML-ready molecular descriptors and featuressubmit_tautomer_search_workflow- Tautomer enumerationsubmit_admet_workflow- ADME/Tox property predictions for drug discoverysubmit_hydrogen_bond_basicity_workflow- Hydrogen bond basicity (pKBHX) predictions
submit_fukui_workflow- Fukui indices for electrophilic/nucleophilic reactivity sites
submit_spin_states_workflow- Spin state energy calculations for different multiplicities
submit_ion_mobility_workflow- Ion mobility mass spectrometry collision cross-section (CCS) predictionssubmit_strain_workflow- Molecular strain energy calculations for rings and cages
submit_double_ended_ts_search_workflow- Transition state search from reactant and product structuressubmit_pose_analysis_md_workflow- Molecular dynamics simulations on docked protein-ligand complexes
submit_docking_workflow- Protein-ligand docking with multiple executables (Vina, QVina2, Smina)submit_batch_docking_workflow- High-throughput docking for virtual screening campaignssubmit_protein_cofolding_workflow- Multi-protein and protein-ligand cofolding predictionssubmit_msa_workflow- Multiple sequence alignment for protein structure prediction
batch_submit_workflow- Submit multiple molecules through any workflow type for high-throughput processing
molecule_lookup- Convert molecule names, CAS numbers, IUPAC names to SMILESbatch_molecule_lookup- Bulk molecule name to SMILES conversionvalidate_smiles- Validate and standardize SMILES strings
create_protein_from_pdb_id- Create protein from PDB ID (e.g., '1HCK')retrieve_protein- Get protein data by UUIDlist_proteins- List all available proteinsupload_protein- Upload custom protein structuresdelete_protein- Remove protein from workspacesanitize_protein- Clean and validate protein structures
retrieve_workflow- Get complete workflow data, status, and results (single source of truth)workflow_stop- Stop running workflowsworkflow_delete- Remove workflows from workspaceretrieve_calculation_molecules- Extract molecular structures from calculationslist_workflows- List all workflows with filtering optionsworkflow_update- Modify workflow parametersworkflow_delete_data- Remove workflow data while keeping metadata
- Python 3.11+
- Rowan API key
- MCP-compatible AI assistant (Claude Desktop, etc.)
Working on a one-click desktop extension (.dxt) for Claude Desktop that eliminates command-line setup! This feature is currently being refined for compatibility with the MCP extension system.
For now, use the standard Package Installation method above, which is fully tested and reliable.
- Documentation: docs.rowansci.com
- or ping me!
If you use this MCP tool in your research, please cite the underlying Rowan platform:
Rowan Scientific. https://www.rowansci.com (accessed 2025-07-01).
For complete citation information including specific computational engines, methods, and workflows used in your calculations, please refer to Rowan's citation guidelines.
To publish a new version to PyPI:
# Update version in pyproject.toml and rowan_mcp/__init__.py
# Build the package
uv build
# Publish to PyPI (requires API token)
uv publish
# Or publish to TestPyPI first
uv publish --index-url https://test.pypi.org/simple/# Start the server first
export ROWAN_API_KEY="your_api_key_here"
uv run python -m rowan_mcp &
# Then inspect
npx @modelcontextprotocol/inspector http://127.0.0.1:6276/sse