BTK CYS481 Covalent Probe Site Validation Certificate AXARA Dataset Engine v3.0
Authors/Creators
Description
BACKGROUND
Three failure modes account for the majority of covalent probe programme losses in kinase drug discovery.
The first is probe site selection error — designing against a residue that is buried, disulfide-bonded, or inaccessible in the relevant biological context. Discovery occurs at the Western blot, typically six to twelve months into the programme.
The second is off-target crosslinking — probe labelling of a kinase other than the intended target, producing a competition signal that is misinterpreted as target engagement. Discovery occurs at the proteomics stage, after the probe has been used to generate biological conclusions.
The third is patient stratification error — running a probe-based assay in a resistance cohort without accounting for which resistance mutations abolish probe labelling and which do not. Discovery occurs when the dataset becomes uninterpretable.
This validation certificate documents the engine output that underpins a commercial data product designed to eliminate all three failure modes before a single probe molecule is synthesised.
WHAT THIS RECORD CONTAINS
This Zenodo record contains the AXARA BTK C481 Validation Certificate v1 — a single PDF documenting the independent validation results of the AXARA Dataset Engine v3.0 applied to Bruton's Tyrosine Kinase CYS481.
The certificate contains: — Engine validation results across four independent BTK crystal structures — Pass/fail determination against the stated criterion — Engine parameters used for the BTK kinase domain configuration — Literature cross-validation table with published experimental evidence — Zenodo DOI field for citation purposes
The certificate does not contain scored residue data, probe scaffold structures, kinome selectivity results, resistance variant classifications, or assay protocols. These are proprietary deliverables available under commercial licence.
VALIDATION RESULTS
Target: BTK — Bruton's Tyrosine Kinase (UniProt Q06187) Primary covalent site: CYS481 — ATP-binding pocket hinge Pass criterion: CYS481 must rank in the top 3 scored residues
5P9J — ibrutinib-bound BTK, 2.2 Å — CYS481 Rank 1 — PASS 6J6M — acalabrutinib-bound BTK, 2.6 Å — CYS481 Rank 1 — PASS 5VFI — WT apo BTK kinase domain, 2.1 Å — CYS481 Rank 1 — PASS 3GEN — WT apo BTK kinase domain — CYS481 Rank 1 — PASS
Result: 4/4 structures PASS. CYS481 ranks first in every structure tested, across both inhibitor-bound and apo forms. No adjustment to engine parameters was made between validation structures.
THE COMMERCIAL DATASET
The validation certificate is the publicly available component of a larger commercial data product. The full BTK C481 Covalent Tractability Map v1 is designed to address each of the three failure modes described above.
Probe site selection error is addressed by the full tractability scoring of all nucleophilic residues in the BTK kinase domain, with evidence tiers assigned to each site and a documented Do Not Pursue list with exclusion reasons.
Off-target crosslinking risk is addressed by a kinome selectivity analysis covering 46 human kinases, identifying which kinases carry a structurally equivalent cysteine with comparable tractability to BTK CYS481.
Patient stratification error is addressed by a resistance variant atlas built from computationally predicted structures of five clinically relevant BTK mutations, classifying each variant for probe labelling compatibility.
Additional deliverables in the full product include a ranked SDA probe scaffold library and four validated assay protocols. Two of the four protocols address experimental contexts for which no published method currently exists.
Pricing and licensing: axara@axara.bio Full product information: axara.bio
ENGINE SUMMARY
AXARA Dataset Engine v3.0 combines three independent scoring signals: binding pocket proximity, solvent accessibility, and SDA chemistry preference. Final score is the geometric mean of all three components. Four exclusion filters are applied: transmembrane topology, disulfide bonds, buried SASA, and fusion protein inserts. Full methodology is documented in the certificate.
The engine has been validated across three target families: KRAS G12C, EGFR C797, and BTK C481. A methods preprint describing the full validation panel is in preparation.