Therapeutics Data Commons (TDC)

Open-science platform curating ML datasets/tasks across the drug discovery pipeline with unified API, splits, and leaderboards.

Kind
meta-platform
Composite
100.0
Benchmarks tracked
83
Direct-linked
99
As of
2026-05-12
Host organisation
Zitnik Lab, Harvard Medical School (+ MIT, Stanford, Georgia Tech collaborators)
Primary contacts
Marinka Zitnik, Kexin Huang, Tianfan Fu
Founded
2021-02
License model
MIT (code); per-dataset licenses for data
Official site
→ site
GitHub
→ GitHub
Count methodology
Scraped tdcommons.ai single_pred/multi_pred/generation overview pages 2026-05-12: single-pred ~38 datasets (ADME/Tox/HTS/QM/Yields/Epitope/Develop/CRISPROutcome), multi-pred ~32 datasets (DTI/DDI/PPI/GDA/DrugRes/DrugSyn/PeptideMHC/AntibodyAff/MTI/Catalyst/TCREpitope/TrialOutcome/ProteinPeptide/PerturbOutcome/scDTI), generation ~13 (MolGen/RetroSyn/Reaction/SBDD). 8 named leaderboard groups.

Rubric

rigor
5
coverage
5
maintenance
5
adoption
5
quality
5
accessibility
5
industry_relevance
5

Breakdown

single_prediction
38
multi_prediction
32
generation
13
leaderboard_groups
8

Notes

Most comprehensive ML-ready therapeutics benchmark hub. NeurIPS 2021 + Nat Chem Bio 2022.

Hosted benchmarks (99)

Direct-linked — each links to the benchmark’s leaderboard / detail page on the host portal.
Benchmark group (8)
Generation (14)
Multi-instance prediction (25)
Single-instance prediction (52)
ADME (absorption)Single-instance predictionADME (distribution)Single-instance predictionADME (metabolism)Single-instance predictionADME (excretion)Single-instance predictionCaco2_WangSingle-instance predictionHIA_HouSingle-instance predictionPgp_BroccatelliSingle-instance predictionBioavailability_MaSingle-instance predictionLipophilicity_AstraZenecaSingle-instance predictionSolubility_AqSolDBSingle-instance predictionHydrationFreeEnergy_FreeSolvSingle-instance predictionBBB_MartinsSingle-instance predictionPPBR_AZSingle-instance predictionVDss_LombardoSingle-instance predictionCYP2C19_VeithSingle-instance predictionCYP2D6_VeithSingle-instance predictionCYP3A4_VeithSingle-instance predictionCYP1A2_VeithSingle-instance predictionCYP2C9_VeithSingle-instance predictionCYP2D6_Substrate_CarbonMangelsSingle-instance predictionCYP3A4_Substrate_CarbonMangelsSingle-instance predictionCYP2C9_Substrate_CarbonMangelsSingle-instance predictionHalf_Life_ObachSingle-instance predictionClearance_Hepatocyte_AZSingle-instance predictionClearance_Microsome_AZSingle-instance predictionLD50_ZhuSingle-instance predictionhERGSingle-instance predictionhERG_KarimSingle-instance predictionAMESSingle-instance predictionDILISingle-instance predictionSkin ReactionSingle-instance predictionCarcinogens_LaguninSingle-instance predictionClinToxSingle-instance predictionTox21Single-instance predictionToxCastSingle-instance predictionHTS_PubChemSingle-instance predictionSARSCoV2_Vitro_TouretSingle-instance predictionSARSCoV2_3CLPro_DiamondSingle-instance predictionHIVSingle-instance predictionQM7bSingle-instance predictionQM8Single-instance predictionQM9Single-instance predictionUSPTO YieldsSingle-instance predictionBuchwald-HartwigSingle-instance predictionSARSCoV2 YieldsSingle-instance predictionIEDB_JespersenSingle-instance predictionPDB_JespersenSingle-instance predictionSAbDab_ChenSingle-instance predictionTAPSingle-instance predictionSAbDab DevelopabilitySingle-instance predictionLeenaySingle-instance predictionTumor_DepMapSingle-instance prediction

← Back to all initiatives

Compare:
Open comparison →