Getting started
Browse, load, or contribute molecular datasets in the LCMD DB
The LCMD DB is a curated collection of computational chemistry datasets — molecules, reactions, and combinatorial fragment libraries — produced by the Laboratory for Computational Molecular Design at EPFL. Each dataset comes with structures, computed properties, and provenance, accessible via a web UI, a Python client, or a REST API.
Pick your path
Browse in the web app
Explore datasets, inspect molecules, and download subsets without writing any code at lcmd-app.epfl.ch.
Load data in Python
Pull datasets into Polars or pandas with the lcmd-db client. Filter,
split, and export.
Contribute a dataset
Define a new subset (molecules, reactions, or fragments) and import it into the platform.
Quick taste
from lcmd_db import load_dataset
data = load_dataset("oscar_nhc")
molecules = data.as_dataset("molecules")
mol = molecules[0]
mol.properties["smiles"]
mol.properties["homo"]See Loading data for filtering, train/test splits, reactions, fragments, and exports.
Reference
The pages on this site cover the platform — what's in it, how to load it, how to contribute. For the canonical client and API references, follow the cards below.
Python client (lcmd-db)
The full Sphinx reference: every class, every method, typed stubs, CLI, assembly templates. The source of truth for the Python client.
REST API
Endpoint reference for direct HTTP access — useful from any language or for ad-hoc curl.
OpenAPI schema
Auto-generated Swagger UI against the live backend.
Stuck?
The FAQ covers common questions. For platform issues, file an issue on GitHub.