OpenCure Explorer

AI Drug Repurposing Dashboard

Multi-pillar computational predictions for 25 underserved diseases · Updated 2026-04-13

587
Candidates
61
Diseases
44
Breakthroughs
363
High Confidence

Confidence: HIGH = multiple evidence types agree; MEDIUM = some support; LOW = computational-only.

Novelty: BREAKTHROUGH = zero published literature; NOVEL = minimal; EMERGING/KNOWN/ESTABLISHED = increasing prior evidence.

Pillars: How many of the 8 independent AI methods support this prediction (more = stronger signal).

MR Score: Mendelian Randomization — causal genetic evidence (0-1, higher = stronger causal support from human genetics).

Score: Combined weighted score across all pillars with convergence bonus. Higher = more AI methods agree.

Click any row to see full evidence: papers, trials, molecular data, and genetic support.

All Predictions

Rank Disease Drug Score Confidence Novelty Pillars MR Evidence

Cross-Disease Drug Network

Drugs predicted as novel candidates for multiple diseases — cross-disease convergence from independent AI pillars suggests shared biological mechanisms and increases prediction confidence.

Top Breakthrough Predictions

Highest-scoring predictions with no prior published evidence — genuinely novel computational discoveries. Click any card for full evidence.

Methodology: 8-Pillar AI Scoring

1

TransE Embeddings

Knowledge graph link prediction using DRKG (5.87M biological relationships)

2

RotatE / PyKEEN

Complex relation pattern scoring via rotation-based graph embeddings

3

TxGNN

Graph neural network for therapeutic use prediction (Harvard)

4

Molecular Fingerprints

RDKit ECFP similarity to known treatments for each disease

5

ChemBERTa

Transformer-learned molecular representations capturing deep structural patterns

6

Gene Signatures

L1000CDS2 transcriptomic reversal — drugs that reverse disease gene expression

7

Network Proximity

STRING protein-protein interaction shortest paths between drug targets and disease genes

8

Mendelian Randomization

Causal genetic evidence from Open Targets — strongest form of drug target validation

Each drug receives a dynamic weighted score across all applicable pillars, with convergence bonuses when multiple independent methods agree. Candidates are then enriched with evidence from PubMed, ClinicalTrials.gov, FDA FAERS, Semantic Scholar, and L1000CDS2.