Score the coincidence
Boltzmann Ensemble covalent ligandability combines pocket fit, cysteine reactivity and sulfur exposure within each conformer, then weights the joint score across the modeled ensemble. This preserves a distinction that separate averages lose: favorable properties occurring together versus appearing in different structures. A calibration procedure evaluates the resulting covalent-ligandability readout within the tested setup.
The conformer is the common reference for those conditions. That changes the shortlist question: which site offers a coherent engagement opportunity, and which only looks attractive after separate scores have been combined?
Different approaches answer different parts of the covalent-site question.
| Question | Familiar approach | BioTwin | Why it matters |
|---|---|---|---|
| Structural opportunity | A single-structure assessment evaluates the supplied geometry; ensemble docking can examine multiple structures. | Scores fit, reactivity and exposure together in each conformer before aggregation. | Distinguishes joint opportunity from favorable properties that occur in different conformations. |
| Biological context | DrugMap measures cysteine ligandability experimentally across cancer cell contexts. | Models structural opportunities and their sensitivity to the chosen ensemble. | The model can nominate hypotheses; contextual engagement data can challenge them. |
| Candidate progression | A site score helps nominate a residue for follow-up. | Can be investigated alongside access geometry and capture kinetics. | A site becomes a specific entry-and-reaction hypothesis, with each added step requiring evidence. |
This is a comparison of scientific roles. No head-to-head advantage over DrugMap or contemporary ensemble-docking methods is claimed.
From a promising state to a capturable site
A useful extension is to pair the ensemble score with molecular fit and access, then reaction-barrier or kinetic analysis. The question becomes whether the ligand can enter a relevant state and react before that opportunity disappears. BioTwin has these component methods; a fully calibrated, experimentally tested composition remains to be established.
Where the evidence holds
Computational calibration studies support site prioritization on small, selected panels. The conformers and Boltzmann weights are model-derived rather than measured equilibrium populations. A target-specific comparison should include a static baseline, matched reactive and unreactive controls, and sensitivity to sampling and weights. Chemical proteomics supplies complementary evidence of engagement in the biological context.