ChEMBL 21: Compound Activity Records Traced to Original Publications

Ye Hu at the University of Bonn assembled two sets of compound activity records from ChEMBL release 21. The data traces compound-target combinations back to their original publications, providing potency measurements and publication references. A separate list of unique publications is included for each set.

Use Cases

• Train models to predict compound-target interactions based on provided potency measurements.
• Link bioactivity data to specific scientific literature for meta-analysis or evidence review.
• Identify unique publications associated with compound-target pairs for bibliometric studies.

Strengths

• Data is sourced from ChEMBL 21, a major public bioactivity database.
• Records are explicitly traced back to original publications, providing provenance.
• The dataset includes two distinct assembled sets for potential comparative analysis.

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

ChEMBL release 21

Collection Method

Traced back to original publications and assembled from ChEMBL.