# Therapeutics Data Commons License: unstated (NOASSERTION); Mdexa hosting: pending. the source says licenses differ per subset and does not list them - Commercial use: Unclear · Derivatives: Unclear · Redistribution: Unclear · File on Mdexa: No - License status: Unstated (NOASSERTION) - Mdexa hosting: Pending — License unstated: the source says licenses differ per subset and does not list them. Link only. - Code repository license: MIT (license-file, verified) - Last verified: 2026-11-14 - Card: https://mdexa.com/datasets/tdc/ ## Key facts - Domain: Pharmacy and drugs, Clinical medicine - Content type: Databases - Size: — - Language: English - Modality: Text, Tabular - Publisher: Zitnik Lab, Harvard - Year: 2020 ## Commercial use of Therapeutics Data Commons - Commercial use: Unclear (Unstated). - License status: Unstated. Verdicts record what sources state and are not legal advice. ## Therapeutics Data Commons license evidence the source says licenses differ per subset and does not list them - **Repository license file** — MIT: GitHub repository license found in snapshot s08 (api-license) — ""spdx_id": "MIT"" — https://github.com/mims-harvard/TDC (2026-11-13; Page opened) - **Terms of use page** — MIT: Statement found in snapshot s05 (readme) — "The TDC codebase is licensed under the MIT license." — https://github.com/mims-harvard/TDC (2026-11-13; Page opened) - **Terms of use page**: Statement found in snapshot s05 (readme) — "For individual dataset usage, please refer to the dataset license on the website." — https://github.com/mims-harvard/TDC (2026-11-13; Page opened) - **Terms of use page** — MIT: Statement found in snapshot s06 (license) — "Permission is hereby granted, free of charge, to any person obtaining a copy" — https://github.com/mims-harvard/TDC (2026-11-13; Page opened) - **Terms of use page** — MIT: Statement found in snapshot s07 (page) — "The TDC codebase is licensed under the MIT license. For individual dataset usage, please refer to the dataset license on the website." — https://github.com/mims-harvard/TDC (2026-11-13; Page opened) ## Source and lineage — ## Known issues with Therapeutics Data Commons - The README defers per-dataset licensing to the website, so the license of the individual datasets is not stated here. - The release is a beta release, so the local copy should be updated regularly. ## Preflight Not run (data not ingested). ## Downloading Therapeutics Data Commons and API access Mdexa does not serve files. Original source: - https://tdcommons.ai/ - https://github.com/mims-harvard/TDC ## Citing Therapeutics Data Commons No recommended citation text. - arXiv:2102.09548 ## Machine-readable - JSON: https://mdexa.com/datasets/tdc.json - 2026-11-14 Known defects, version notes and access statements from cited sources (papers, release notes, official pages) were added; each shows who said it, when, and what kind of source it is. - 2026-11-14 Practical fields (scope, identifiers, access methods, versions, known defects) were added by the automated pipeline from stored page snapshots; every statement carries a verbatim quote. - 2026-11-14 Known defects, version notes and access statements from cited sources (papers, release notes, official pages) were added; each shows who said it, when, and what kind of source it is. - 2026-11-14 Known defects, version notes and access statements from cited sources (papers, release notes, official pages) were added; each shows who said it, when, and what kind of source it is. - 2026-11-14 Re-adjudicated under the current rules; all machine checks still pass. - 2026-11-14 Known defects, version notes and access statements from cited sources (papers, release notes, official pages) were added; each shows who said it, when, and what kind of source it is. - 2026-11-14 Known defects, version notes and access statements from cited sources (papers, release notes, official pages) were added; each shows who said it, when, and what kind of source it is. - 2026-11-14 Re-adjudicated under the current rules; all machine checks still pass. - 2026-11-14 Known defects, version notes and access statements from cited sources (papers, release notes, official pages) were added; each shows who said it, when, and what kind of source it is. - 2026-11-13 Added via the automated pipeline and passed all checks (gates, verbatim quotes, rule engine, two independent extractions). ## Practical information ### Content scope - The text states that TDC includes 66 AI-ready datasets spread across 22 learning tasks and spanning the discovery and development of safe and effective medicines. > To date, TDC includes 66 AI-ready datasets spread across 22 learning tasks and spanning the discovery and development of safe and effective medicines. https://arxiv.org/abs/2102.09548 (2026-11-13) - The text states that TDC covers learning tasks including target discovery, activity screening, efficacy, safety and manufacturing across biomedical products such as small molecules, antibodies and vaccines. > TDC covers a wide range of learning tasks, including target discovery, activity screening, efficacy, safety, and manufacturing across biomedical products, including small molecules, antibodies, and vaccines. https://tdcommons.ai/ (2026-11-13) ### Identifiers - The text shows a dataset retrieved by the name HIA_Hou from the ADME therapeutic learning task under single-instance prediction. > For a concrete example, to obtain the HIA dataset from the ADME therapeutic learning task in the single-instance prediction problem: https://raw.githubusercontent.com/mims-harvard/TDC/HEAD/README.md (2026-11-13) ### Access - Web interface only: https://github.com/mims-harvard/TDC ### Versions - TDC is versioned as the PyTDC Python package; the latest PyPI release is 1.1.15, uploaded 2025-03-31. No separate data-release numbering was found. > "upload_time":"2025-03-31T15:20:14" https://pypi.org/pypi/PyTDC/json (2026-11-13) - The text states that TDC is regularly updated with new datasets and tasks, such as antibody therapeutics and gene editing. > TDC is regularly updated with novel datasets and tasks, such as antibody therapeutics and gene editing. https://tdcommons.ai/overview/ (2026-11-13) ### Known defects and misuse - The author writes that TDC is "equally flawed" as MoleculeNet, judged by the criteria he sets out in the article. > other widely used benchmarks like the Therapeutic Data Commons (TDC) are equally flawed. https://practicalcheminformatics.blogspot.com/2023/08/we-need-better-benchmarks-for-machine.html (2026-11-13) - Walters: the MoleculeNet BBB set contains 59 duplicate structures, 10 of which carry conflicting labels, plus 11 SMILES RDKit cannot parse. > First, the dataset contains 59 duplicate structures. This is not the sort of thing we want to have in a benchmark. But wait, it gets worse, 10 of these duplicate structures have different labels. https://practicalcheminformatics.blogspot.com/2023/08/we-need-better-benchmarks-for-machine.html (2026-11-13) - Walters: in the MoleculeNet BACE set 71% of molecules have at least one undefined stereocenter. > 71% of the molecules have at least one undefined stereocenter, 222 molecules have 3 undefined stereocenters, and one molecule has a whopping 12 undefined stereocenters. https://practicalcheminformatics.blogspot.com/2023/08/we-need-better-benchmarks-for-machine.html (2026-11-13) ### Suited for - The text states that TDC is a resource to access and evaluate AI methods and supports the development of AI methods for drug discovery applications. > Therapeutics Data Commons is a resource to access and evaluate AI methods, supporting the development of AI methods, with a strong bent towards establishing the foundation of which AI methods are most suitable for drug discovery applications and why. https://tdcommons.ai/ (2026-11-13) - TDC authors state their experiments show even the strongest algorithms fall short on distribution shift and generalisation to novel data. > demonstrating that even the strongest algorithms fall short of solving key therapeutics challenges, including real dataset distributional shifts, multi-scale modeling of heterogeneous data, and robust generalization to novel data points. https://arxiv.org/abs/2102.09548 (2026-11-13) - Walters: MoleculeNet physiology datasets (BBB, toxicology) have complex endpoints that make them less than ideal for benchmark comparisons. > The complexity of these endpoints makes the data less than ideal for benchmark comparisons. https://practicalcheminformatics.blogspot.com/2023/08/we-need-better-benchmarks-for-machine.html (2026-11-13) ### No sourced statement yet Breaking changes, Unsuited tasks, Commercial licensing ## Continue in this database This record is one of 109 processed and verified records (datasets, data resources and open models). To find others, call `https://mcp.mdexa.com/v1/datasets?q=` or connect the MCP server at `https://mcp.mdexa.com/mcp`; no key is needed for 20 calls per minute. Rows are under 1 KB; an empty result means "not listed". For a list of items, one batch check (free account) returns a report. Guide: /agents.md. Similar records: TopU-LBVS: Multi-Target Benchmark for Ligand-Based Virtual Screening (https://mdexa.com/datasets/topu-lbvs/); BindingDB (https://mdexa.com/datasets/bindingdb/); DrugCentral (https://mdexa.com/datasets/drugcentral/).