A computational docking workflow with a deliberately narrow interpretation
The receptor is an AlphaFold ABL1 monomer, not an experimental imatinib-bound structure or a fusion-specific BCR–ABL complex. The output ranks poses within this setup and does not validate binding.
Can an automated workflow generate plausible imatinib poses in an ABL1 pocket?
Imatinib targets the ABL kinase domain clinically, but this session used the AlphaFold model for UniProt P00519. It did not model the BCR fusion partner or the experimentally observed inactive DFG-out conformation.
Within a fixed AlphaFold ABL1 receptor and selected fpocket site, what poses and Vina scores are generated for imatinib, and how do they overlap known contact residues?
| Ligand | Imatinib; PubChem CID 5291 |
|---|---|
| Receptor | AlphaFold ABL1, UniProt P00519 |
| Pocket selection | fpocket with residue-overlap review |
| Docking | AutoDock Vina 1.1.2; exhaustiveness 32; 10 poses |
| Search box | 25 Å × 25 Å × 25 Å |
Compound retrieval, ADMET summary, structure audit, pocket selection, and docking
Ten poses scored from −8.8 to −9.4 kcal/mol in the selected pocket
The ten Vina poses ranged from −8.8 to −9.4 kcal/mol.
The selected fpocket site overlapped 12 of 39 reference contact residues.
Imatinib passed four of five listed drug-likeness filters; the Ghose filter failed.
Low solubility, moderate CYP/hERG concern, and QED 0.389 were reported.
The receptor and docking protocol do not reproduce the clinical binding state
- The receptor is AlphaFold ABL1, not a fusion-specific BCR–ABL complex.
- The model is not an experimental imatinib-bound DFG-out conformation.
- Rigid-receptor docking does not sample induced fit or important protein flexibility.
- The selected pocket only partially overlaps the reference contact set.
- Vina scores are heuristic and cannot establish affinity, selectivity, residence time, or potency.
- The ADMET summary is rule based and requires experimental or validated predictive follow-up.
This is a pose-generation and workflow demonstration against an ABL1 model. It is not validation of imatinib’s known mechanism and is not a drug-design recommendation.
Artifacts recorded in the Pipette session
imatinib_pubchem.csvCompound identifiers and propertiesimatinib_admet_summary.csvRule-based property flagsAF-P00519-F1.pdbAlphaFold ABL1 receptor modelfpocket_results.csvCandidate pocket scoresselected_pocket.jsonSelected coordinates and residue overlapposes_all.sdfTen exported docking poses