liuganghuggingface/moldata-clintox

ClinTox clinical toxicity and FDA approval

ClinTox clinical toxicity and FDA approval. DeepChem/MoleculeNet release.

Part of the Molecular Datasets collection.

Get started

from datasets import load_dataset
dataset = load_dataset("liuganghuggingface/moldata-clintox", "default")

Where the data comes from

• Original data: published with the original study
• Original study: A Data-Driven Approach to Predicting Successes and Failures of Clinical Trials
• Redistributed by: MoleculeNet (DeepChem) (dataset page)

Tables

Config	Splits	Rows	Targets
default	full	1,480	FDA_APPROVED FDA APPROVED (classification; binary), CT_TOX Clinical-trial toxicity (ClinTox) (classification; binary)

data/ contains rows with readable molecular structures; identical observations are collapsed. observations/ contains all source rows.

Source label columns are text, including values such as <10. SMILES_original retains the input representation; SMILES_canonical provides a canonical representation. structure_key groups molecular components, and structure_status records parsing outcomes.

Data quality

• experiment: Labels are measured in a lab.
• many sources: Compiled from several labs or papers, whose protocols differ. Independent IC50 values for the same compound and target typically differ by a factor of about 5 (Kalliokoski et al. 2013).
• 0.2% unreadable: 4 rows of 1,484 have a SMILES that RDKit cannot read; they are left out of the processed table.
• 29.7% undefined stereo: 29.7% of molecules have a stereocenter whose configuration is not given, so mirror-image forms cannot be told apart.
• 3 of 52 stereoisomer sets differ: Found 52 groups of molecules that differ only in stereochemistry. In 3, the values are 10-fold apart (log-scale labels), more than one standard deviation apart, or in different classes. That can be a real effect or a mistake in the source; a model that ignores stereochemistry treats these molecules as the same.
• 5.6% PAINS alerts: 5.6% of molecules contain a PAINS substructure, which often looks active in screening assays without real activity.
• 100% conflicting repeats: In a typical task, 100% of molecules listed more than once have different classes in different rows.
• 7.5% in rarer class: In a typical task, 7.5% of labeled molecules are in the rarer class. Accuracy is misleading here; use precision and recall.
• 100% measured per task: In a typical task, 100% of molecules have a value. A missing value means not measured, not inactive.
• 1,461 molecules per task: A typical task has 1,461 labeled molecules.
• no source split: Make your own split. A random split puts near-identical molecules in both training and test sets, so results look better than they will on new compounds. Splitting by scaffold, or by date when dates are given, gives a more realistic estimate.

Dataset notes

• FDA approval is a regulatory outcome rather than a direct toxicity assay. CT_TOX is sparse and class-imbalanced; overlaps TDC ClinTox.

License

Dataset-wide license: unspecified. Original source.

Citation

Cite the original study and data source:

• Gayvert et al. A Data-Driven Approach to Predicting Successes and Failures of Clinical Trials (2016)
• Wu et al. MoleculeNet: a benchmark for molecular machine learning (2018)
• Official DeepChem loader and task definitions