Rowan

You use the Rowan MCP Server to run Rowan’s molecular design and simulation tools through a configurable MCP gateway. This server lets you connect an MCP client to Rowan’s computation engines, run complex workflows, and retrieve results from a centralized, scriptable interface.

How to use

Connect your MCP client to Rowan MCP Server using the HTTP/SSE endpoint or run the local server to operate directly from your machine. You can ask the integrated AI assistant to perform chemistry calculations, optimize structures, predict properties, and run various workflows. Start with a simple property prediction or geometry optimization and progressively combine tools for multi-step workflows.

How to install

Prerequisites you need before installation are Python 3.11+ and a MCP-compatible client. You will also need an API key for Rowan’s services.

{
  "mcpServers": {
    "rowan": {
      "type": "http",
      "url": "http://127.0.0.1:6276/sse"
    }
  }
}

Option 2 lets you install Rowan MCP Server locally and then start it. After installing, configure the MCP client to point at the local HTTP endpoint.

export ROWAN_API_KEY="your_api_key_here"
rowan-mcp

Additional notes

Security: keep your API key private and set it in your environment before starting the server. Make sure the port and host are accessible from your MCP client. If you run the local server, you can start it with the standard start command shown above and then connect via the provided HTTP endpoint.

Configuration and usage tips

You typically use the HTTP endpoint for remote access or the local stdio flow for a full debugging or development setup. When using the HTTP method, your MCP client should point to the SSE endpoint at http://127.0.0.1:6276/sse. When using the local flow, start the server with uv and the Python module to run Rowan’s MCP interface.

Security and access control

Protect your API key and restrict access to the server. Use network controls to limit who can reach http://127.0.0.1:6276/sse and ensure the host machine is secured. Rotate API keys as needed and monitor access logs from your MCP client.

Available tools

submit_basic_calculation_workflow

Submit a workflow to perform energy calculations, optimizations, and frequency analyses using multiple engines (omol25, xtb, psi4) to obtain optimized geometries and related data.

submit_conformer_search_workflow

Run conformer searches with multiple modes (rapid, careful, meticulous) to explore low-energy structures.

submit_conformers_workflow

Generate and enumerate molecular conformers beyond the initial search to broaden structural coverage.

submit_multistage_opt_workflow

Execute multi-stage geometry optimizations with sequential accuracy levels for refined structures.

submit_scan_workflow

Perform molecular scans (dihedral, bond, angle) with wavefront propagation to map energy landscapes.

submit_irc_workflow

Compute intrinsic reaction coordinates for transition states to analyze reaction pathways.

submit_pka_workflow

Calculate microscopic pKa values with customizable pH ranges and element selections.

submit_macropka_workflow

Compute macroscopic pKa values across pH and charge ranges.

submit_solubility_workflow

Predict solubility across solvents and temperatures.

submit_redox_potential_workflow

Predict electrochemical reduction/oxidation potentials.

submit_descriptors_workflow

Generate ML-ready molecular descriptors and features.

submit_tautomer_search_workflow

Enumerate tautomeric forms.

submit_admet_workflow

Predict ADME/Tox properties for drug discovery.

submit_hydrogen_bond_basicity_workflow

Predict hydrogen bond basicity (pKBHX) of molecules.

submit_fukui_workflow

Compute Fukui indices to identify electrophilic and nucleophilic reactive sites.

submit_spin_states_workflow

Calculate spin state energies for different multiplicities.

submit_ion_mobility_workflow

Predict ion mobility collision cross-sections (CCS) for molecules.

submit_strain_workflow

Calculate molecular strain energies for rings and cages.

submit_double_ended_ts_search_workflow

Search for transition states from reactant and product structures.

submit_pose_analysis_md_workflow

Run MD simulations on docked protein-ligand complexes.

submit_docking_workflow

Docking workflows against proteins using multiple engines (Vina, QVina2, Smina).

submit_batch_docking_workflow

High-throughput docking for virtual screening campaigns.

submit_protein_cofolding_workflow

Predict multi-protein and protein-ligand cofolding scenarios.

submit_msa_workflow

Perform multiple sequence alignments 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, and IUPAC names to SMILES.

batch_molecule_lookup

Bulk conversion of molecule names to SMILES.

validate_smiles

Validate and standardize SMILES strings.

create_protein_from_pdb_id

Create a protein model from a PDB ID (e.g., 1HCK).

retrieve_protein

Fetch protein data by UUID.

list_proteins

List all available proteins in your workspace.

upload_protein

Upload custom protein structures.

delete_protein

Remove a protein from your workspace.

sanitize_protein

Clean and validate protein structures.

retrieve_workflow

Get complete workflow data, status, and results from a single source of truth.

workflow_stop

Stop a running workflow.

workflow_delete

Remove a workflow from the workspace.

retrieve_calculation_molecules

Extract molecular structures from calculations.

list_workflows

List all workflows with filtering options.

workflow_update

Modify workflow parameters.

workflow_delete_data

Remove workflow data while preserving metadata.