Multi-Omics Target Discovery for Rheumatoid Arthritis

Autoimmune

Multi-Omics Target Discovery for Rheumatoid Arthritis

Pharma discovery group needed a ranked, druggable target list with evidence dossiers before Q4 portfolio gate—replacing a stalled GWAS-only internal analysis.

Company

Mid-size pharma discovery group (internal RA franchise)

Timeline

October 2025

Engagement

Multi-Omics Target Discovery Pipeline

Multi-Omics Target Discovery

3 wks: Computational delivery
847: Targets evaluated
12: Priority targets + dossiers
100%: Portfolio gate approval

The Challenge

A pharma discovery group was launching a new rheumatoid arthritis franchise and needed druggable targets with strong genetic, pathway, and single-cell evidence. Internal bioinformatics had produced 200+ GWAS hits but lacked druggability scoring, synovial cell-type specificity, and competitive landscape context. Leadership required a decision-ready target list with full evidence dossiers before Q4 portfolio gate review in four weeks.

Business Constraints

Budget: $125K (exploratory target ID ceiling)
Timeline: Ranked dossiers in 3 weeks, gate deck in week 4
Must include druggability, safety tier, and competitive landscape per target
Reconcile with internal GWAS list without discarding prior work

TargetIQ Approach

Week 1: Multi-Omics Integration and Knowledge-Graph Scoring

Input

Public RA datasets: GWAS (Okada et al., 2014), synovial tissue RNA-seq (12 cohorts, n=412)
Single-cell RNA-seq from RA synovium (6 published atlases, 284,000 cells)
Client internal GWAS shortlist (218 genes)
Open Targets and DisGeNET disease–gene associations

Methods

Cross-cohort differential expression meta-analysis (847 unique genes)
Knowledge-graph GNN propagation over Reactome + STRING networks
Single-cell cell-type specificity scoring (fibroblast, macrophage, T-cell subsets)
Genetic colocalization (eQTL + GWAS integration, 47 loci tested)

Output

847-gene ranked list by composite disease relevance score
Pathway map linking top 50 genes to RA inflammatory cascade (JAK-STAT, TNF, B-cell)
Reconciliation report: 34/218 internal hits upgraded, 89 downgraded with rationale

Week 2: Druggability, Safety, and Competitive Landscape

Druggability scoring (PDB structure availability, AlphaFold pocket analysis, known ligand classes, modality fit) filtered 847 genes to 94 tractable targets. Safety assessment (DepMap essentiality, knockout phenotypes, GTEx tissue breadth) removed 31 high-liability targets including 4 from the internal shortlist. Competitive landscape analysis (ClinicalTrials.gov, Cortellis, patent filings) flagged 18 crowded targets. Knowledge-graph centrality re-weighted remaining 63 targets across four priority tiers. Evidence confidence tiers assigned: concordant (multi-source), discordant (conflicting layers), single-source.

Week 3: Evidence Dossiers and Portfolio Gate Package

Output

Top 12 priority targets with full evidence dossiers (genetic, expression, pathway, druggability, safety, competition)
36-target expanded tier list with 2–3 backups per priority band
Recommended validation experiments per target (CRISPR knockdown, biomarker assay, patient stratification)
Competitive landscape brief per Priority A target (active trials, approved drugs, white-space assessment)
Portfolio gate presentation deck with decision framework and budget allocation recommendations

Priority Target List with Evidence Dossiers

Top 12 targets delivered with full dossiers; 36 targets in expanded tier list. Dossier template includes genetic evidence table, pathway context, druggability rationale, and gate recommendation.

Ranked RA targets by disease relevance, druggability, safety, and evidence concordance
Rank	Target	Druggability	Evidence Tier	Safety Tier	Gate Status
1	TYK2 (JAK family)	0.92	Concordant	Moderate	Approved — Priority A
2	GPR65	0.78	Concordant	Low risk	Approved — Priority A
3	CD19 (B-cell)	0.95	Concordant	Moderate	Approved — Priority A (crowded IP flagged)
4	IL6ST (gp130)	0.88	Concordant	Moderate	Approved — Priority A
5	PTPN22	0.71	Concordant	Low risk	Approved — Priority B (novel)
6	IRF5	0.74	Discordant	Low risk	Approved — Priority B
7–12	Mixed (incl. TNFRSF1A, STAT4, BLK)	0.65–0.85	Moderate to concordant	Low to moderate	8/12 approved at gate

Results and impact

Speed, validation, and business outcomes

Speed vs. Internal Bioinformatics Approach

Metric	Internal Bioinformatics	TargetIQ	Improvement
Timeline	4 to 6 months	3 weeks	6× faster
Cost	$400K+ (FTE time)	$125K	69% savings
Targets Evaluated	218 (GWAS only)	847 (multi-omics + GNN)	4× broader evaluation
Decision-Ready Output	Gene list only	12 dossiers + gate deck	Portfolio-ready deliverable
Safety Pre-Filter	Not performed	31 unsafe targets removed	Avoided downstream liability

Portfolio Gate Outcomes (4 weeks post-delivery)

Investment committee review and literature validation on top 12 targets.

Target	Druggability	Clinical Evidence	Gate Outcome	Notes
TYK2	0.92	Deucravacitinib approved (PsA); RA expanding	Funded — 2026 budget	Strong precedent; dossier drove unanimous approval
GPR65	0.78	Preclinical only; strong GWAS + scRNA	Funded — exploratory	Novel GPCR; white-space confirmed in dossier
CD19	0.95	Multiple CAR-T trials	Deferred — crowded IP	Competitive brief prevented duplicate spend
IL6ST	0.88	Tocilizumab approved (RA)	Funded — differentiated modality	Novel biologic angle identified in dossier
PTPN22	0.71	GWAS only; undrugged	Funded — Priority B	High genetic confidence; structure-enabled flag
Rank 6–12	0.65–0.85	Mixed	5/7 approved	2 deferred pending additional cohort data

12/12: Dossiers accepted by review committee
9/12: Targets with prior clinical evidence
3: Novel undrugged targets identified
31: Unsafe targets filtered pre-gate

Immediate Wins

Portfolio gate approved: 9 of 12 priority targets received 2026 budget allocation; 3 deferred with clear rationale
Competitive advantage: GPR65 and PTPN22 identified as white-space targets with no competing Phase 2+ programs
Team alignment: shared evidence dossiers eliminated 6 weeks of internal debate on target prioritization
Internal list reconciliation: 89 internal GWAS hits correctly downgraded with documented rationale, preserving team trust

Strategic Advantages

Single-cell weighting surfaced GPR65—missed entirely by bulk tissue and GWAS-only internal analysis
Knowledge-graph GNN identified IRF5 pathway centrality not captured by differential expression alone
Safety pre-filter prevented pursuit of 31 targets with essential-gene or broad tissue expression liability
Gate deck format adopted as internal template for future franchise reviews

Follow-on engagement
Q1 2026: small molecule discovery engagement on TYK2 and GPR65 via HelixForge. Estimated combined cost: $650K. Target: ranked chemical matter within 4 weeks per target.

Model validation

Lessons and recommendations

What Worked

Single-cell RNA-seq weighting improved target ranking vs. bulk tissue analysis alone (Spearman ρ +0.34 vs. internal list)
Genetic colocalization (eQTL + GWAS) was strongest predictor of portfolio committee approval (9/9 concordant targets funded)
Evidence confidence tiers (concordant/discordant/single-source) resolved internal disputes on conflicting targets within 48 hours
Reconciliation report format preserved internal bioinformatics credibility while upgrading methodology

Challenges and Mitigations

GPR65 scored lower on druggability (0.78) due to limited structural data for this GPCR class.

Mitigation: Flagged as structure-enabled target with AlphaFold model quality assessment (pLDDT 82); recommended cryo-EM partnership in dossier validation plan.

Two targets (IRF5, BLK) had conflicting evidence between GWAS significance and single-cell expression specificity.

Mitigation: Assigned discordant evidence tier with explicit resolution pathway: request client proprietary synovial biopsy cohort for tie-breaker analysis.

When to use TargetIQ for autoimmune target discovery

New franchise or indication expansion requiring ranked target lists with gate-ready dossiers
Internal bioinformatics producing GWAS or expression lists without druggability or safety context
Portfolio gate or investment committee deadlines within 4 to 8 weeks
Need for single-cell, genetic, and pathway evidence integrated into one decision framework

ROI: approximately 8:1 (FTE cost avoidance + 6× faster timeline + avoided pursuit of 31 unsafe targets + 6 weeks debate time saved).

Next steps: advance top 2 to 3 gated targets into chemistry or biologics pipelines; optionally ingest proprietary patient cohort for discordant-tier resolution.

About This Engagement

Client profile: Mid-size pharma, internal RA franchise, 12-person discovery team
Project duration: 3 weeks (computational delivery) + 4 weeks (gate review)
Total cost: $125K
Date: October 2025

This case study is anonymized at client request. Target names are published biology; program-specific strategy details have been redacted. Full dossiers available under NDA.

Run a similar program with TargetIQ.

Tell us about your target. We will scope a 2 to 4 week pilot.

Schedule a Consultation [email protected]