Illustrative structural analysis
Ask which observation rules out an interface.
Start with the public LINGO-1–Li81 Fab structure, PDB 4OQT. A retained structural-analysis test uses its deposited arrangement and a grid of rotations and translations. It checks each candidate against three conditions: avoid severe clashes, retain enough atom contacts and place the known epitope at the interface. The output is designed to retain admissible arrangements and explain the first constraint that rejects each alternative.
- 4OQT + alternativesSupply candidate arrangements
Deposited pose plus an explicit set of rotations and lateral shifts.
- 3 constraintsApply the same checks
Clash separation, contact count and epitope coverage.
- Survivors + reasonsPlan a discriminating observation
Separate poses that fail the epitope check from those excluded by geometry.
- Make the experimental information count · Compare the full constraint set with steric and contact checks alone. The useful check is whether the epitope condition excludes additional alternatives. A filter that only removes clashes has not established the proposed interface.
- Carry the reason into the next experiment · For an unsolved complex, use independently supplied epitope or mutagenesis evidence. If several arrangements survive, choose a new observation that distinguishes them. HADDOCK and other integrative methods also use experimental restraints; the useful artifact here is the candidate set tied to explicit exclusion reasons.
Narrow the interface, one observation at a time
Steric compatibility alone can leave many arrangements standing. Mutagenesis or a blocking epitope may distinguish them. The useful result is the surviving hypothesis together with the observation that excluded an alternative.
Information-driven docking tools such as HADDOCK already use experimental information. BioTwin’s interface work addresses a narrower job: checking candidate arrangements or deposited contacts under the constraints supported by the particular operation.
| Question | Familiar approach | BioTwin focus | Practical consequence |
|---|---|---|---|
| Starting point | HADDOCK can generate and refine complex models using structural and experimental information. | Begin with the candidate arrangements or deposited structures accepted by the selected method. | Choose the operation that matches the structural information you actually have. |
| Question | Which complex models are consistent with the docking inputs and restraints? | Which proposed arrangements survive the supported contact, steric, mutagenesis or blocking constraints? | Learn which observation rules a candidate in or out. |
| Next evidence | Model confidence still depends on the quality and relevance of the restraints. | Surviving arrangements remain hypotheses, with unsupported interface assumptions visible. | Choose a discriminating interaction experiment. |
Choose the method according to the available inputs. Surviving BioTwin interface hypotheses still require experimental validation.
Match the question to the structural input
Candidate-pose operations and static or deposited-contact analysis have different input requirements. This work is constraint-guided analysis, not a general promise to predict an interface from any pair of unbound proteins. A candidate that survives the checks still needs binding or structural evidence.
Use the surviving arrangement to ask a harder design question
A proposed interface could inform recruiter geometry, a linked binding element or an intervention at a modified protein. Those follow-up methods need the relevant state and compatible structural inputs.
- Recruiter mechanisms · Investigate whether an arrangement could support productive recruitment. Recruiter mechanisms
- Multi-head binding · Check what a proposed interface implies for a linked architecture. Multi-head binding
- Modified proteins · Revisit assumptions when charge or modification state changes. Modified proteins