PAINS Compounds: Promiscuous vs. Dark Chemical Matter with Prediction Models

Swarit Jasial from the University of Bonn provides data sets of promiscuous PAINS (PROM_PAINS) and dark chemical matter PAINS (DCM_PAINS). Support vector machine models built on original and balanced training data are included. The dataset is available under an Open Access (green) license.

Use Cases

• Training machine learning models to distinguish between promiscuous and dark chemical matter compounds.
• Validating compound screening filters based on PAINS substructures.
• Analyzing chemical property patterns associated with promiscuous binding behavior.

Strengths

• Includes support vector machine models built on both original and balanced training data.
• Provides distinct data sets for two important compound classes (PROM_PAINS and DCM_PAINS).

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Last update date is unknown; freshness unverified.

Provenance

Source

University of Bonn