AlphaFold3 Covalent Ligand Discovery for Kinase Inhibitors

Encompassing results from a prospective virtual screening campaign using AlphaFold3 (AF3) cofolding predictions to discover covalent small molecule inhibitors, specifically targeting the model kinase BTK. It demonstrates AF3's application for ranking true covalent binders over property-matched decoys and includes validation via co-crystallography showing subangstrom accuracy. The dataset is provided in XLSX format and is 33.5 MB in size.

Use Cases

• Analyze AF3 cofolding predictions and associated confidence metrics to rank covalent binders for kinase targets.
• Validate predicted binding modes from virtual screening campaigns using co-crystallography data showing subangstrom accuracy.
• Compare the performance of AF3-based screening against state-of-the-art covalent docking tools for protein kinases.
• Investigate the chemical distinctness and proteomic selectivity of discovered covalent small molecules like the BTK inhibitor.

Strengths

• Includes experimental validation via co-crystallography showing subangstrom accuracy of predicted AF3 binding modes.
• Demonstrates a practical application of AlphaFold3 for novel covalent ligand discovery against kinases, a prolific drug target family.
• Dataset is 33.5 MB, making it a manageable size for detailed analysis.

Limitations

• The dataset size is 33.5 MB, indicating it is a small-scale study, not a large-scale screening library.
• Specific row and column counts are unknown, limiting precise assessment of data volume and feature scope.
• Focus is primarily on protein kinases, particularly BTK, which may limit generalizability to other protein families.

Provenance

Source

figshare

Collection Method

Generated from a prospective virtual screening campaign and experimental validation study using AlphaFold3.

Freshness

Last updated on 2026-03-20.