Innovator›Discovery Campaign

Target → governed compute → persisted evidence → reviewable shortlist

Design the campaign around the scientific decision—not the software.

Qualify the target, choose a transparent screening funnel, run real structure-based computation, persist post-docking ADMET/novelty/synthesis evidence, inspect the evidence and continue the same candidates into a persistent Discovery Programme. Docking and model/search outputs remain computational or retrieved evidence; experiments and qualified human review remain the decision gate.

Your target firstReal VinaADMET + novelty + synthesisFail-closed readinessPersistent handoff

Programme name

Disease / indication

Discovery objective Minimum 10 characters

DEFAULTYour target / PDBUse the biological target selected for your programme. Production compute remains blocked until readiness is approved. EXAMPLETry a qualified exampleUse the current NDM-1 example profile to experience the managed workflow.

Target-specific by designPocket, receptor state, cofactors, pose rules, ensemble and optional physics are governed per target.

Biological target

PDB ID / structure identifier

Qualified example profile

Core readiness contract

A custom target is not made “ready” by entering a PDB ID. Required structural and protocol checks remain explicit.

Not checked

Readiness has not been checked.

Advanced physics options

RECOMMENDEDAuto when qualifiedUse only target protocols already qualified for the selected system. OffComplete the core campaign without MD/MM-GBSA. Request qualificationFlag a separate target-specific physics qualification request when appropriate.

Runtime details: OpenMM and AmberTools are implementation runtimes, not client-facing scientific choices. The campaign executes only a pinned, qualified protocol; advanced physics never becomes mandatory simply because software is installed.

CLIENT PILOTVerified 100Exactly 100 generated → 100 fast real Vina → 30 accurate → 10 deep → up to 5 for integrated evidence + human review. Pilot 1K1,000 generated → 1,000 fast real Vina → 100 accurate → 30 deep. Focused 20K20,000 generated → 2,000 fast → 500 accurate → 50 deep. Advanced 100K100,000 generated → 5,000 fast → 1,000 accurate → 100 deep.

Core compute—engineering estimate

Allocation—configurable internal basis

Review shortlist—computational candidates

Physics—optional / target-dependent

Scientific boundary: approving this campaign authorizes computational and database-retrieval work only. The final shortlist remains hypotheses requiring medicinal-chemistry, experimental and human/CRO review.

Campaign summary

One place to see the next blocker.

Planning

Programme—

Target—

Structure—

Core readinessNot checked

Campaign—

PhysicsAuto when qualified

Complete the programme and target fields.

Recent campaigns

Persistent account history; resume on another browser or device.

Evidence contract

The campaign ranks hypotheses; it does not manufacture certainty.

CCOMPUTEDDocking, rescoring, ADMET/safety prediction, synthesis route search and optional physics outputs.Model

RRETRIEVEDPubChem exact-identity evidence; not patentability or freedom-to-operate proof.Database

HHuman/CRO gateShortlist progression requires qualified review and experimental confirmation.Required

Campaign control room

Campaign running

Loading campaign state…

———

Running

Current scientific activity

Quantitative progress appears only when the executor reports real counts/percentages. Other stages use discrete states.

Resolving current stage…

Campaign artifacts

Downloads appear when the relevant stage has produced an artifact. Integrated CSV/JSON and the verified PDF are derived from persisted evidence, not browser state.

No campaign artifacts yet.

Future downstream capability

FEP is intentionally not an active campaign stage.
It may become appropriate later for an experimentally established congeneric analogue series. It is not shown as “complete” during screening.

Computational evidence package ready for review

Shortlist ready for persistent scientific review.

The final candidates remain computational hypotheses. BayesPharma preserves docking and post-docking evidence in the persistent Programme ledger so the next work happens on Candidate Passports rather than in an isolated result table.

Continue to Discovery Programme →

Human/CRO-review shortlist

Structure-first cards expose the ranking channels without pretending one score proves binding.

0 candidates

Loading shortlist…

Research-use computational decision support. Target qualification is target-specific. Optional MD/MM-GBSA is executed only when scientifically qualified. ADMET/safety predictions, PubChem identity retrieval and AiZynthFinder routes do not replace experimental or legal validation.

Plan Discovery Campaign

Core readiness contract

Campaign running

Current scientific activity

Campaign artifacts

Future downstream capability

Shortlist ready for persistent scientific review.

Human/CRO-review shortlist