Developing breakthrough therapeutics across Metabolic, Oncology, and Rare Disease — powered by the #1 global ranking in AI-drug discovery patent families for 2025. The lead biased GLP-1R program is in active development toward functional validation.
Patent Leadership in AI Drug Discovery
The Benchmark That Matters
Deep EigenMatics holds the #1 global ranking in AI-drug discovery methods patent families for 2025, establishing an intellectual property position that restructures competitive dynamics in computational pharmaceutical development. The company holds 24 patent filings — 10 granted or allowed (7 issued U.S. patents; 3 allowed and en route to issuance), 14 pending — protecting foundational mathematical architectures. All patents are assigned 100% to Deep EigenMatics, Inc.; inventor Stephen G. Odaibo. Founded February 2025, Deep EigenMatics has reached seven issued U.S. patents and a #1 global 2025 ranking — an IP-generation velocity without obvious precedent among AI-drug-discovery startups.
The company's AI-drug discovery patent families granted in 2025 exceeded the entire Big Pharma sector by 2x and surpassed all major AI drug discovery competitors — including Google DeepMind (Isomorphic Labs), Insilico Medicine, and Recursion Pharmaceuticals. Methodology is verifiable via the linked PR Newswire release citing public USPTO data.
#1
Global AI-Drug Discovery Methods Patent Rank — 2025, across all pharmaceutical entities
2x Big Pharma
AI-drug discovery methods patent families granted in 2025 exceeded the entire Big Pharma sector
Platform IP
Foundational mathematical architectures, not individual compounds
Exit Optionality
Acquisition, partnership, or IPO pathways
Press & Recognition
Deep EigenMatics Ranks #1 Globally in AI Drug Discovery, Outpacing All of Big Pharma in 2025 Patent Output
February 3, 2026 - Source: PR Newswire
Deep EigenMatics, Inc., a pioneer in high-velocity Artificial Intelligence for drug design, announced today that it has secured the #1 global ranking for new U.S. patent families in AI Drug Discovery methods for 2025.
2025 U.S. Patent Count: AI Drug Discovery Innovation
This chart shows Deep EigenMatics' leading position in AI drug discovery patent families compared to all major competitors and Big Pharma in 2025.
The chart reflects 2025 AI-drug discovery methods patent families by entity, based on public USPTO data as cited in the PR Newswire release. Deep EigenMatics ranked #1 globally, ahead of all AI drug discovery companies and the entire Big Pharma sector.
Lead Asset: Oral GLP-1R Program
The global GLP-1 receptor agonist market represents a $100B+ opportunity long dominated by injectables, and the first oral agents have now reached market: oral semaglutide (Rybelsus, 2019; oral weight-loss formulation approved December 2025) and orforglipron (Foundayo, Eli Lilly) — the first oral small-molecule GLP-1 receptor agonist, FDA-approved April 1, 2026. These first-generation orals validate the oral thesis but are optimized primarily for receptor affinity. Deep EigenMatics is engineering a next-generation, signaling-biased oral GLP-1R agonist — separating Gs signaling from beta-arrestin recruitment to design for the therapeutic window, not affinity alone.
DEEP DISCOVERY™ AI Platform
Deep EigenMatics is utilizing its proprietary DEEP DISCOVERY™ AI Platform — integrating patented conditional multi-headed neural networks, probabilistic reasoning architectures, CAM-guided transformer methodology, and dynamic context refresh mechanisms — to engineer a next-generation oral GLP-1 candidate. DEEP DISCOVERY learns structure-function correlation across the GPCR superfamily in a telescoping manner, integrating structural, ligand, and functional data in tandem, and applies that learned intelligence to generate candidate ligands with targeted safety and efficacy profiles. The lead biased GLP-1R program is in active development toward functional validation; seed-stage and pre-clinical. IND-enabling studies are targeted to begin within approximately six months of lead compound synthesis.
IP Coverage
The candidate program will be protected by IP coverage across process, manufacturing methods, composition of matter, and formulation strategy — building a defensible position around a differentiated, signaling-biased oral GLP-1 candidate.
Oncology Program: EVICT CANCER
Deep EigenMatics' EVICT CANCER program deploys DEEP DISCOVERY to design novel CXCR4 antagonists targeting the cancer stem cell sanctuary mechanism — the primary driver of therapy resistance and relapse in solid and hematologic malignancies. EVICT CANCER leverages DEEP DISCOVERY's AI platform — applied with precision to CXCR4 — to engineer novel antagonists with the exact profile required for clinical impact in GBM and multiple myeloma.
Primary Indication: Glioblastoma Multiforme (GBM)
No approved therapy has been designed from the outset for direct anti-tumor activity, oral bioavailability, and blood-brain barrier penetrance simultaneously. EVICT CANCER targets all three by design.
Secondary Indication: Multiple Myeloma (MM)
CXCR4/CXCL12 axis blockade disrupts the bone marrow niche that protects malignant plasma cells from standard-of-care therapy. EVICT CANCER targets this sanctuary mechanism to sensitize resistant disease to existing treatment regimens.
DEEP DISCOVERY Platform: Three-Program Pipeline
Deep EigenMatics advances three independently motivated programs sharing the DEEP DISCOVERY platform — each targeting a distinct therapeutic area with its own scientific rationale and development timeline.
GLP-1R Metabolic Program
Lead program. In active compute. Lead biased GLP-1R program in active development toward functional validation; seed-stage and pre-clinical. IND-enabling studies targeted to begin within approximately six months of lead compound synthesis.
EVICT CANCER Oncology Program (CXCR4)
Primary indication: Glioblastoma multiforme (GBM). Secondary indication: Multiple myeloma (MM). Novel CXCR4 antagonists targeting the cancer stem cell sanctuary mechanism. In development.
Rare Disease Program
Target to be announced. Pipeline expansion stage. Leveraging DEEP DISCOVERY's therapeutic-agnostic architecture to address high-unmet-need rare disease indications.
AI-Driven Economics: Redefining Drug Development
Deep EigenMatics leverages AI to restructure the economics, timeline, and risk profile of pharmaceutical innovation compared to traditional empirical approaches.
Dramatically Reduced Capital Requirements
Computational design enables multi-program development at a fraction of traditional pharmaceutical investment.
Accelerated Development Timelines
In-silico validation compresses the path from discovery to IND-ready candidate by years compared to empirical methods.
De-Risked Clinical Progression
Pathway prediction and safety screening reduce clinical uncertainty before entering costly human trials.
Multi-Program Platform
Therapeutic-agnostic discovery architecture enables simultaneous advancement across Metabolic, Oncology, and Rare Disease programs.
IP Protection
Foundational mathematical architectures combined with composition and formulation patents create a defensible innovation position.
The transition from capital-intensive, empirical drug development to AI-driven computational design represents a categorical shift in how pharmaceutical value is created.
Foundational Mathematics: The Engine
Deep EigenMatics' competitive advantage originates from a portfolio of granted and pending patents protecting mathematical innovations that advance the state of computational drug discovery. Each capability below is protected by granted U.S. patents. Full patent numbers, claim charts, and freedom-to-operate analyses available under NDA.
Seven Granted U.S. Patents Protecting Core Platform
Conditional Multi-Headed Neural Networks
Proprietary architecture for joint synthesis of amino acid sequences and structural parameters in a single synchronized process, enabling discovery of novel compositions with optimal target-binding affinity and enhanced drug-likeness.
Probabilistic Reasoning for Autoregressive Models
Methodology for diverse library generation that comprehensively explores chemical space to identify high-affinity candidates with superior drug-likeness constraints.
Recursive Transformers for Protein-Protein Interaction
Architecture for accurately mapping complex protein-protein interactions and simulating downstream signaling cascades, enabling in-silico safety screening and efficacy prediction. Granted intellectual property protects this architecture.
CAM-Guided Transformer Methodology
Architecture for learning structure-function correlation from large-scale data spanning the GPCR superfamily and beyond, in a telescoping manner — integrating structural, ligand, and functional data in tandem — and applying that learned intelligence to a specific receptor to generate candidate ligands with desired safety and efficacy profiles.
Dynamic Context Refresh Mechanisms
System that dynamically updates context during drug generation, ensuring molecules continuously adhere to drug-likeness constraints and evolve towards optimized properties.
Early Fusion of Natural and Protein Language Models for Generative AI-Based Protein and Drug Design
US 12,651,642, granted June 9, 2026. Joint early-fusion of natural-language and protein-language models for generative protein and drug design.
Retrieval-Augmented Fusion Language Models for AI-Based Protein and Drug Design
US 12,633,423 B2, granted May 19, 2026. Retrieval-augmented architecture conditioning generative protein/drug design on retrieved sequence and structural context.
Patent Portfolio: 24 Filings — 7 Issued U.S. Patents, 3 Allowed, 14 Pending
Deep EigenMatics holds 24 patent filings — 10 granted or allowed (7 issued U.S. patents; 3 allowed and en route to issuance), 14 pending — constituting the foundational intellectual property architecture of the platform. The seven granted patents below establish discrete and enforceable positions in AI-driven drug discovery. All patents are assigned 100% to Deep EigenMatics, Inc.; inventor Stephen G. Odaibo.
Leadership Infrastructure: Execution Pedigree
Dr. Stephen G. Odaibo, M.D., M.S. (Math), M.S. (Comp. Sci.) — Founder & CEO
- Fellowship-trained in Medical Retina, Ocular Oncology, and Uveitis, Kellogg Eye Center, University of Michigan
- Trained in the Nobel Prize-winning Lefkowitz Lab (Duke University) — birthplace of GPCR science and foundational biology of GLP-1 receptors (2012 Nobel Prize in Chemistry)
- Inventor of the platform's foundational patents; prosecuted pro se
- Author: The Foundational Mathematics of Artificial Intelligence
- Dual expertise: deep mathematical foundations and clinical drug development insight
Advisory
Andrew Hill — Data Infrastructure & Intellectual Property Strategy
20-year veteran in data, distributed systems, and cloud infrastructure. Named inventor on foundational patents spanning geospatial systems, logistics optimization, and large-scale data architectures. Experienced in complex IP enforcement and defense matters involving major technology incumbents. Technology founder and operator with successful exits. Provides strategic guidance on computational scaling, capital structuring, and long-term IP positioning.
The GLP-1 Market: Scale, Trajectory, and the Oral Opportunity
A Market Structurally Awaiting Oral Delivery
The global GLP-1 receptor agonist market is among the fastest-expanding therapeutic categories in modern pharmaceutical history. Current injectable market leaders — semaglutide, tirzepatide — have demonstrated transformative clinical efficacy, yet their delivery modality imposes a structural ceiling on patient penetration.
Conservative epidemiological modeling indicates that needle-aversion and administration burden exclude an estimated 40–60% of the clinically eligible obesity and type 2 diabetes population from sustained GLP-1 therapy. The first oral agents are now reaching market and beginning to expand access — but they are optimized for affinity, leaving signaling quality and receptor selectivity as the open frontier.
Deep EigenMatics is advancing a patent-protected, AI-native platform toward a biased oral GLP-1R agonist — engineered for signaling selectivity (Gs over beta-arrestin) rather than affinity alone — at a moment when institutional capital is seeking exactly this exposure.
Raising Seed Capital to Achieve IND-Ready Status
Deep EigenMatics is raising seed capital to advance its lead oral GLP-1R program toward functional validation and IND-enabling studies, while expanding the discovery platform across multiple therapeutic verticals.
Capital Deployment Strategy
Lead Asset Advancement
Digital candidate lock, in-vitro validation, in-vivo studies, and IND-enabling studies
Platform Expansion
Metabolic, Oncology, and Rare Disease programs across multiple therapeutic verticals
IP Generation
Continued IP generation and defensive patent portfolio expansion
Team & Infrastructure
Scientific and operational team expansion to support program advancement.
Value Creation Milestones
Computational Candidate Lock
AI-validated lead candidate selected and locked from the DEEP DISCOVERY platform, with a full in-silico data package — prior to synthesis.
Experimental Validation
In-vitro and in-vivo data confirming efficacy and safety profile of lead asset
IND-Ready Status
Full IND submission readiness achieved, enabling Phase I initiation
Detailed capital allocation and financial projections available under NDA.
Strategic Positioning: Request NDA Access
Deep EigenMatics operates at the intersection of AI and pharmaceutical development. The combination of foundational IP, capital-efficient execution, and a multi-program platform architecture creates durable strategic optionality.
Why This Matters
The foundational patent portfolio supports valuation resilience across exit scenarios. Any institution seeking to compete in AI-accelerated drug development must either:
License Deep EigenMatics' Foundational IP
The fastest and most capital-efficient path to AI-driven drug discovery
Multi-Decade Alternative Development
Capital-intensive efforts to develop alternative mathematical frameworks — a barrier that grows with each additional patent filing.
Access the Full Data Room
The following are available to qualified investors under NDA:
Technical Specifications
Detailed technical documentation and complete patent portfolio analysis
Financial Models
Granular financial models and milestone-based valuation framework
Leadership & Commitments
Advisory relationships and program details.
Deep EigenMatics is not a single-asset biotech. The foundational IP portfolio positions the company as infrastructure for the next generation of pharmaceutical discovery.
Deep EigenMatics: AI-Native Drug Discovery Across the GPCR Superfamily