Skip to content

chematic

Kentaro Tanabe (kent-tokyo)

chematic is a Rust cheminformatics toolkit with Python and WebAssembly bindings for local molecular analysis, plus an optional MCP server. RDKit compatibility covers a documented subset.

Catalog updated ·

Overview

chematic provides a shared Rust chemistry core for native applications, Python notebooks and browser tools. It turns molecular representations into structures, calculated descriptors and fingerprints, with similarity and substructure operations for exploring compound collections. The source is available under MIT OR Apache-2.0. The optional chematic-mcp component connects chemistry tools to MCP clients; the library itself is not an autonomous research agent.

Compatibility is defined per operation and binding. The RDKit-style layer covers selected APIs, and default Morgan fingerprints are not bit-identical to RDKit. 3D generation, pKa/ADMET screening, IUPAC naming, Markush/polymer expansion and rich CDXML editing remain experimental or bounded. Calculated outputs do not establish experimental efficacy or safety. Results should be assessed on the intended chemistry and exact software artifact.

Key Features

  • Parse and write supported SMILES/SMARTS and common MOL/SDF molecular records.
  • Calculate descriptors and fingerprints; perform Tanimoto similarity and substructure searches.
  • Use documented Rust, Python, Node and WebAssembly interfaces backed by the shared Rust core.
  • Run supported browser chemistry locally in WebAssembly without a chematic backend.
  • Expose local chemistry tools through chematic-mcp using stdio or optional self-hosted Streamable HTTP.
  • Return explicit failures for unsupported cases; canonical_smiles() is not a universal identity key.

Use Cases

  • Explore compound collections in a Python notebook and prepare descriptor or fingerprint features for downstream machine learning.
  • Embed molecular parsing, property cards and similarity exploration in a browser application.
  • Let an MCP client request documented chemistry operations, while keeping experimental screening outputs subject to domain review.

How to Use

The documentation offers pip install chematic for supported Python wheel platforms and npm install @kent-tokyo/chematic for browser integration. A minimal Python example is:

import chematic
mol = chematic.from_smiles("CC(=O)Oc1ccccc1C(=O)O")
print(mol.mw, mol.logp, mol.tpsa)

Browser integrations initialize the WASM module before parsing molecules and release handles afterward. Node requires explicit WASM byte loading. For MCP, configure the client to launch chematic-mcp with an empty argument list for stdio; select the package version using its release documentation. The HTTP option is self-hosted and supplies no authentication or OAuth. The optional name_to_smiles tool sends names to PubChem. See the installation guide and MCP README for setup and limits.

Sources and verification

This entry describes capabilities documented by the linked project sources. Catalog edits are not evidence of a successful run, compatibility test or scientific performance evaluation.

Read the editorial and verification policy

Related resources

AQME

Open Source

AQME builds quantum-mechanical working environments end to end: conformer generation, input preparation, output checking and descriptor calculation across its CSEARCH, CMIN, QPREP, QCORR and QDESCP modules.

Open sourcePython

Cheminformatics · Computational Chemistry · Quantum Chemistry

OpenChemistry libraries behind Avogadro, providing molecular data handling, rendering and extensible chemistry functionality.

Open sourceC++

Cheminformatics · Computational Chemistry

ChemCP is an MCP App that turns SMILES strings into interactive 2D molecular diagrams and computed descriptors using RDKit.js, with an in-chat viewer for hosts that support MCP Apps.

TypeScriptJavaScript

Cheminformatics

ChemCrow is a Python chemistry agent package built with Langchain that connects language-model reasoning to RDKit, paper-qa, chemical databases, and reaction-planning tools.

Open sourcePython

Cheminformatics · Literature & Research

Open-source MCP server for local chemical data cleaning, CAS format and checksum validation, duplicate detection, and dataset quality checks through AI assistants.

Open sourceTypeScript

Cheminformatics · Scientific Data

Chemistry Development Kit (CDK) is a Java library for chemical representations, file conversion, structure searching, fingerprints, QSAR descriptors and molecular rendering within other programs.

Open source

Cheminformatics

Related guides

Overview

MCP vs Skills vs Agents vs APIs: What's the Difference?

Understand how APIs, MCP servers, Skills, and agents divide interface, procedural guidance, execution, and planning responsibilities in chemistry workflows—and how to choose or combine them without confusing access with scientific validity.