← Datasets & benchmarks

GenesisPotency-2026

GenesisPotency-2026 is a datasets & benchmarks tool whose code is published by Genesis-Molecular-AI on GitHub, first public in 2026. Genesis Molecular AI's audited protein–ligand potency training set: 5.41M rows (pIC50/pKi/pKd or active/inactive) from ChEMBL, BindingDB and PubChem with row-level provenance, plus potency-ranking and hit-ID benchmarks.

Visit website ↗

CategoryDatasets & benchmarks
Developer (GitHub)Genesis-Molecular-AI
Websitehuggingface.co/datasets/genesisml/GenesisPotency
GitHubGenesis-Molecular-AI/genesis-potency-dataset
Released2026-11-20 (source)
LicenseCC-BY-SA-4.0 (Dataset CC BY-SA 4.0 (GitHub API and HF card); adapts ChEMBL (CC BY-SA 3.0), BindingDB (CC BY 4.0 / CC BY-SA 3.0 for ChEMBL-derived), PubChem and UniProt (CC BY 4.0); benchmark configuration adapts OpenFE/FEP+ (MIT), CACHE (CC BY 4.0), OpenBind (CC0) and CASP16 (no licence stated). Keep NOTICE.md attributions. Ruby model weights/code are not released)
What this allowsUnknown
No confirmed terms from a primary source; ask the authors before using it in a product. What do these mean?
License sourceapi.github.com/repos/Genesis-Molecular-AI/genesis-potency-dataset · checked 2026-11-21
PricingOpen source
Tags#binding-affinity#potency-prediction#protein-ligand#benchmark#chembl
Last checked · links verified

Can I use GenesisPotency-2026 commercially?

Unclear. We could not confirm the terms from a primary source; check with the authors before commercial use.

What the license types mean · not legal advice.

Specs

Size5,410,616 rows (1,773,271 potency measurements; 3,637,345 binary calls incl. 200,662 synthetic decoys); 1,038,258 compounds, 3,640 protein sequences, 69,939 assay contexts; 49 columns
Modality—
AccessSingle Parquet file on Hugging Face (genesisml/GenesisPotency, revision v2026; ~11 GB RAM to load fully); docs/examples on GitHub
GitHub stars11
Interface—

Compare with…

Pick 1–3 other Datasets & benchmarks tools to compare side by side with GenesisPotency-2026.

Select 1–3 tools

Related

OMol25

Meta FAIR's open dataset of DFT calculations on diverse molecules, with evaluations and baseline models.

Datasets & benchmarksOpen sourceUnknown#dft

ML-ready datasets and benchmarks for therapeutics.

Datasets & benchmarksOpen sourcePermissive (commercial OK)#benchmark

ProteinGym

Benchmarks for protein fitness and mutation effect prediction.

Datasets & benchmarksOpen sourcePermissive (commercial OK)#benchmark