W
Wilbur Salt OS
Enterprise Cognition Engine

Indication Expansion Methodology

A systematic, evidence-bounded approach powered by thousands of synthetic coworkers—each mapped to a defined domain role. De-risk transformation. Simulate before you execute.

Pro Tip

This methodology powers the "Drug Seeking Indication Expansion" workflow in the R&D module. Select any drug from your portfolio to generate ranked expansion hypotheses backed by real-world evidence.

Core Principles

Three foundational principles guide every analysis in the Wilbur SaltOS enterprise cognition engine.

Evidence-Bounded
All hypotheses are grounded in verifiable data sources—clinical trials, peer-reviewed literature, and regulatory databases. No speculative claims without supporting evidence.
Phenotype-Centric
Analysis focuses on patient phenotypes—observable characteristics, comorbidities, biomarkers, and treatment patterns—capturing biological complexity beyond mechanism-of-action alone.
Systematic Ranking
Hypotheses are ranked using multi-dimensional scoring that integrates phenotypic similarity, co-occurrence, market opportunity, and regulatory pathway feasibility.

7-Stage Workflow

Click on each stage to explore the detailed methodology and outputs.

1

Comparative PICO Evidence

Analyze up to 10 comparators using the PICO framework (Population, Intervention, Comparison, Outcome).

  • Identifies relevant comparators from same therapeutic class
  • Ranks drugs by efficacy metrics (RRR, NNT, ORR)
  • Validates head-to-head trial status and superiority claims
  • Ensures outcomes align with actual clinical trial endpoints
2

Regulatory Context

Drug-specific intelligence on approval history and regulatory status.

3

RWD Intelligence

Real-world evidence from HTA bodies and clinical guidelines.

4

Market Intelligence

Commercial context for prioritizing expansion opportunities.

5

Clinical Trials

Active and completed studies for drug and key comparators.

6

Phenotype Clustering

Patient cluster analysis based on phenotype similarity.

7

Hypothesis Report

Ranked recommendations with confidence scores and evidence.

Connected Data Sources

Live API connections to clinical, regulatory, and scientific databases power the analysis engine.

ClinicalTrials.gov
Live
Global clinical trial registry

450,000+ studies

PubMed/MEDLINE
Live
Biomedical literature database

35M+ citations

FDA DailyMed
Live
Drug label database

140,000+ labels

FDA Press Releases
Live
Regulatory news feed

Real-time RSS

EMA News
Live
European regulatory updates

Real-time RSS

NIH News
Live
Research updates

Real-time RSS

Phenotype Similarity Analysis

Patient populations are characterized by observable features across five categories. Similarity between indications is calculated using weighted feature overlap, capturing the biological and clinical complexity that determines therapeutic response.

Diagnosis Patterns

25% weight
Primary diagnosisComorbiditiesDisease stage

Laboratory Results

25% weight
BiomarkersGenetic testsImaging

Prescription History

20% weight
Treatment sequencesDrug classesAdherence

Procedures

15% weight
Surgical interventionsDiagnostic procedures

Demographics

15% weight
Age distributionGenderGeography
Confidence Scoring
Each hypothesis receives a composite confidence score based on multiple factors.
Phenotype Similarity40%
Co-occurrence Score40%
Biomarker Overlap10%
Mechanism Alignment10%

Confidence Tiers

High ≥60%Strong evidence, accelerated development
Medium 40-59%Moderate evidence, validation needed
Low <40%Preliminary signal, further investigation

Ready to Explore Indication Expansion?

Select a drug from your portfolio and generate evidence-backed expansion hypotheses in minutes.