MODULE 05 · Discovery

Drug repurposing engine

Focus on one disease area and build an evidence graph for existing medicines. Prioritise hypotheses using transparent scoring and contradictory evidence before considering expensive experiments.

THE FIRST USEFUL PRODUCT

Start with a bounded problem.

Who it is for
Translational researchers and nonprofit teams investigating underfunded treatment questions.
First product
A disease-specific evidence map with ranked, cited and explicitly uncertain repurposing hypotheses.
Next decision
The scope and evidence standard are agreed before ranking begins.
Product boundary

A ranking is not a treatment recommendation. Existing approval for one indication does not establish safety or effectiveness for another.

PRODUCT 0 → 1.0

The Phase 0–5 strategy.

Each phase has a concrete deliverable and an acceptance gate. These are proposed development plans, not completed scientific or product validations.

PHASE 0 / 0% MILESTONECurrent planning

Bound the disease question

Define what a useful repurposing hypothesis must contain.

Work to do

  • Choose one disease area with an expert reviewer.
  • Document current treatment context and the research gap.
  • Set inclusion criteria for medicines and evidence.

Acceptance gate

The scope and evidence standard are agreed before ranking begins.

Publish: Disease brief and hypothesis rubric.

PHASE 1 / 20% MILESTONEPlanned

Build the evidence graph

Connect entities while preserving where each relationship came from.

Work to do

  • Map drug, target and disease identifiers across sources.
  • Attach evidence type, date and source to every edge.
  • Keep negative, conflicting and indirect evidence distinguishable.

Acceptance gate

Sampled edges can be traced to sources and mapping errors are quantified.

Publish: Versioned evidence graph and mapping report.

PHASE 2 / 40% MILESTONEPlanned

Create transparent rankings

Make the first prioritisation model understandable.

Work to do

  • Start with a documented rule-based scoring baseline.
  • Display which evidence raises or lowers each score.
  • Generate a cited hypothesis card with missing-evidence fields.

Acceptance gate

The same snapshot and weights reproduce every score and explanation.

Publish: Ranking prototype and hypothesis cards.

PHASE 3 / 60% MILESTONEPlanned

Test for hindsight bias

Check the method without leaking future knowledge into past predictions.

Work to do

  • Use time-separated evidence snapshots where feasible.
  • Compare with simple popularity and evidence-count baselines.
  • Test whether small weight changes radically change rankings.

Acceptance gate

The evaluation reports leakage controls, baseline comparisons and unstable hypotheses.

Publish: Retrospective benchmark and sensitivity report.

PHASE 4 / 80% MILESTONEPlanned

Review with specialists

Learn whether the outputs support a useful research decision.

Work to do

  • Ask experts to review a preselected hypothesis set.
  • Record disagreement, plausibility concerns and evidence gaps.
  • Define follow-up experiments only with qualified research partners.

Acceptance gate

Each prioritised hypothesis has a recorded expert rationale and a feasible next evidence step.

Publish: Expert review report and follow-up research plan.

PHASE 5 / 100% MILESTONEPlanned

Release the hypothesis workspace

Keep the evidence and ranking history inspectable over time.

Work to do

  • Publish snapshot versions and ranking-change explanations.
  • Support review records and reproducible exports.
  • Monitor source updates and preserve superseded results.

Acceptance gate

Independent review can reproduce the graph and rankings without implying treatment readiness.

Publish: Product 1.0 repurposing explorer.

Progress stays at 0% until the starting scope is accepted and subsequent milestone evidence is published. Phase 1–5 targets are 20%, 40%, 60%, 80% and 100%. These percentages track development, not treatment effectiveness.

WHAT THIS DEPENDS ON

Build with the right foundations.

  • Datasets & models library: Open Targets, ChEMBL, UniProt and trial records.
  • Clinical evidence module: traceable extraction and source review.
  • Disease specialist: hypothesis review and experimental relevance.

Partner roles above are requirements. No partner participation or endorsement is claimed.

FUTURE INTERACTIVE PRODUCT

Repurposing evidence explorer

Open after identifier mapping, evidence provenance and ranking reproducibility are independently checked.

  1. Choose a disease evidence map.
  2. Inspect drug-target-disease connections and contradictions.
  3. Adjust documented research-priority weights.
  4. Export cited hypotheses for expert review.
View the portal specification
PUBLIC CHANGELOG

What changed. What is still planned.

Drug repurposing engine

Phase 0–5 strategy and portal plan published

Published the module strategy, concrete phase deliverables, acceptance criteria and future portal workflow. This is a planning update; product completion remains 0%.

Entries are published with site updates. Planned work is labelled separately from completed product milestones.