Selectivity Profiles of 10,060 Multi-Kinase Inhibitors Across 141 Human Kinases

10,060 multi-kinase inhibitors and their selectivity profiles across 141 human kinases, derived from ChEMBL23 high-confidence data. The dataset reports 596 protein kinase pairs, each with UniProt IDs, shared inhibitors, and pIC50 potency values. It was compiled by Filip Miljković at the University of Bonn.

Use Cases

• Assessing compound selectivity based on reported pIC50 potency differences
• Identifying shared inhibitors for kinase pairs to study polypharmacology
• Benchmarking computational models for kinase inhibitor binding predictions
• Prioritizing compounds for experimental validation based on selectivity profiles

Strengths

• 10,060 compounds provide a substantial chemical space for analysis
• Potency data includes pIC50 values for 141 distinct human kinase targets
• Data is sourced from the high-confidence subset of the ChEMBL23 database

Limitations

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

Provenance

Source

ChEMBL23 high-confidence data

Collection Method

Reported list compiled from the database