# Plex Research > Plex is the scientific intelligence platform for drug discovery. Our mission is to increase the world's capacity to discover cures by harnessing massive amounts of biomedical data. Plex queries structured biomedical databases — genomics, proteomics, compound activity, clinical trials — to deliver evidence-backed answers grounded in real data, not LLM guesswork. Trusted by 40+ pharma and biotech discovery teams. ## About Plex Research was founded in 2017 by Douglas Selinger and is headquartered in Cambridge, Massachusetts (625 Massachusetts Ave, Central Square — halfway between Harvard and MIT). The company is led by CEO Brian Gilman and CSO Douglas Selinger. Plex has served 40+ pharmaceutical and biotech companies. The platform integrates billions of experimental data points across multi-omics, chemical biology, and scientific literature. ## How Plex Is Different Plex is an LLM for data and databases used in biomedical research — not a literature search tool. While other AI tools search published text and generate summaries from language models, Plex queries structured biological and chemical datasets directly. Every answer is traceable to specific data sources and records, with reproducible, evidence-scored outputs ready for R&D decisions. Key differences from literature-based AI: - Built on billions of structured data points from biomedical databases - Queries genomic, proteomic, clinical, and chemical datasets directly - Every answer traceable to specific data sources and records - Integrated chemical intelligence tooling for compound analysis - Reproducible, evidence-scored outputs ready for R&D decisions ## Core Technology ### Focal Graph The Focal Graph is Plex's proprietary biomedical knowledge graph. It uses centrality algorithms — inspired by Google's PageRank — to identify the strongest connections across billions of biomedical data points. Given a scientific query (e.g., "Which targets can overcome EGFR inhibitor resistance in NSCLC?"), the Focal Graph builds a subgraph around the query entity, runs PageRank to identify important nodes, and generates structured output with evidence-backed results. The Focal Graph achieved 81.2% top-1 accuracy in target identification benchmarks. ### AutoPlex AutoPlex is Plex's autonomous AI research agent. It plans and executes complex drug discovery research programs, interprets results, and provides full details of its strategy and the experimental validation data uncovered. AutoPlex can run large-scale research campaigns across the entire knowledge graph autonomously. ### LLM Integration Plex combines the Focal Graph with large language models to create an AI that understands chemical structures and omics data. Scientists ask questions in natural language and receive data-driven answers with explicit links to the supporting experimental evidence — not just text summaries, but traceable results from real datasets. ## Data Coverage Plex searches across 60+ databases and billions of relationships, including: - Transcriptomics (gene expression profiles across tissues and conditions) - Proteomics (protein expression and interaction data) - Metabolomics (metabolite profiles and pathways) - Genomics (genetic variants, GWAS, and functional genomics) - Chemical structures and small molecule bioactivities - Clinical studies and trial data - Scientific literature (PubMed and beyond) - Customer proprietary data (securely integrated) ## Example Queries and Answers **Q: Which targets are implicated in EGFR-resistant NSCLC?** A: Key resistance targets include MET amplification, KRAS G12C, and HER2/ERBB2 — with secondary mutations in EGFR T790M and C797S driving acquired resistance to first- and third-generation TKIs. Plex integrates genomic, proteomic, and Perturb-seq signatures to rank targets by evidence strength. **Q: What biomarkers predict response to anti-PD1 therapy in melanoma?** A: Tumor mutational burden (TMB), PD-L1 (CD274) expression, and interferon-gamma gene signatures are the strongest predictors. Plex cross-references clinical trial outcomes with transcriptomic profiles to surface emerging biomarker candidates like CXCL9 and LAG3 co-expression patterns. **Q: Evaluate TREM2 as a therapeutic target for Alzheimer's disease.** A: TREM2 has strong genetic validation — the R47H loss-of-function variant confers ~2–4x increased AD risk. It plays a central role in microglial phagocytosis and neuroinflammation. Druggability is challenging given its extracellular immunoglobulin domain, with antibody agonists currently in Phase I/II trials. ## Products & Services ### Scientific Intelligence Platform (For Scientists) Scientists ask questions in natural language and receive data-driven answers with explicit links to supporting experimental evidence. The platform searches billions of data points across multi-omics, chemical biology, and literature in seconds, with every result traceable to experimental evidence. Use cases include: - Target identification and validation - Biomarker discovery - Mechanism of action elucidation - Safety signal detection - Competitive intelligence - Indication expansion analysis ### Program Evaluation (For Executives) Structured program evaluation that unifies biological evidence, internal data, therapeutic feasibility, and risk into one repeatable assessment. Evaluations are delivered in weeks, not months. The framework covers 7 integrated evaluation dimensions: 1. Mechanistic biology and target-disease association 2. Human genetics evidence 3. Chemistry feasibility and druggability 4. Safety and toxicity signals 5. Internal proprietary data concordance 6. Translational readiness and biomarker strategy 7. Competitive landscape and differentiation ### Developer API Programmatic access to the Plex biomedical knowledge graph. Three API calls from raw input to biomedical insight: resolve entities, search the graph, and run autonomous research campaigns with AutoPlex. Built for computational biologists, bioinformaticians, and data scientists. ### Free Pipeline Intelligence Report Personalized AI-generated reports showing where a drug pipeline has hidden upside, overlooked targets, and mechanistic blind spots. Available at no cost. ### Enterprise Pipeline Intelligence Report Structured, mechanism-aware analysis across multi-omic datasets for large pharma R&D organizations. Identifies differentiated targets, clarifies biological pathways, and reduces translational risk. ## How It Works 1. Ask a question — Pose any scientific question, from target identification to biomarker discovery. Plex understands both biological and chemical context. 2. Plex searches the knowledge graph — Billions of data points spanning biology and chemistry are integrated and scored in real time. 3. Get structured, actionable answers — Receive evidence-backed results with confidence scores, source citations, and built-in chemical intelligence tooling. ## Scientific Outputs - Target Evaluation Reports: Structured assessment across 7 dimensions — biology, genetics, druggability, safety, clinical signal, competitive landscape, and proprietary data concordance. - Biomarker Panels: Prioritized candidate biomarkers with predictive power estimates, validation status, and recommended assay strategies. - Competitive Intelligence: Real-time landscape analysis of competing programs, MOA overlap, clinical trial status, and differentiation opportunities. ## Use Cases - **Target Identification** — Leverage the Focal Graph's centrality algorithms to identify and validate novel drug targets across billions of biomedical data points. - **Biomarker Discovery** — Discover clinically relevant biomarkers by integrating transcriptomics, proteomics, metabolomics, and genomics data. - **Target Evaluation** — Score and compare targets across 7 evidence dimensions before committing capital. - **Translational Science** — Bridge the gap between preclinical findings and clinical outcomes with AI-driven translational readiness assessments. - **Mechanism of Action Elucidation** — Uncover how drugs and compounds exert their effects by mapping molecular interactions and pathway perturbations. ## Team - **Brian Gilman** — Chief Executive Officer. LinkedIn: https://www.linkedin.com/in/briangilman/ - **Douglas Selinger** — Chief Scientific Officer & Founder. LinkedIn: https://www.linkedin.com/in/douglasselinger/ - **Timothy Wall** — Chief Architect - **Jed Gaetz** — Principal Scientist - **Ehab Khalil** — Principal Scientist - **Eleni Stylianou** — Principal Scientist - **Oren Levy** — Biology Consultant ## Pages - https://plexresearch.com/ — Homepage: Scientific intelligence platform for drug discovery - https://plexresearch.com/researcher.html — Scientist landing page: AI-powered drug discovery platform for biologists and chemists - https://plexresearch.com/developer.html — Developer API: programmatic access to the biomedical knowledge graph - https://plexresearch.com/academic.html — Academic collaborations: AI-driven target deconvolution and multi-omics analysis - https://plexresearch.com/plex-live.html — Try Plex Live: interactive demo of the scientific intelligence platform - https://plexresearch.com/executive.html — For Executives: structured program evaluation for drug discovery governance - https://plexresearch.com/plex-report.html — Free pipeline intelligence report request - https://plexresearch.com/plex-report-large-pharma.html — Enterprise pipeline intelligence report for large pharma R&D - https://plexresearch.com/target-identification.html — Use case: AI-powered target identification - https://plexresearch.com/target-evaluation.html — Use case: Evidence-scored target evaluation across 7 dimensions - https://plexresearch.com/biomarker-discovery.html — Use case: AI-driven biomarker discovery across multi-omics data - https://plexresearch.com/translational-science.html — Use case: Translational science powered by the Focal Graph - https://plexresearch.com/moa-elucidation.html — Use case: Mechanism of action elucidation - https://plexresearch.com/team.html — Team page with bios and leadership - https://plexresearch.com/news.html — Latest news, conferences, and press - https://plexresearch.com/contact.html — Contact form and office location - https://plexresearch.com/blog.html — Blog: scientific insights and platform updates ## Contact Website: https://plexresearch.com Email: info@plexresearch.com LinkedIn: https://www.linkedin.com/company/plex-research Address: 625 Massachusetts Ave., 2nd Floor, Cambridge, MA 02139 ## Frequently Asked Questions **What is Plex Research?** Plex is a scientific intelligence platform for drug discovery. Unlike literature-based AI tools, Plex queries structured biomedical databases — genomics, proteomics, compound activity, and clinical trials — to deliver evidence-backed answers with full source traceability. It is built on a knowledge graph spanning billions of data points from 60+ databases. **What is the Focal Graph?** The Focal Graph is Plex's proprietary knowledge graph that uses centrality algorithms — inspired by Google's PageRank — to identify the strongest connections across billions of biomedical data points. It distills vast, noisy datasets into concise, transparent, data-driven hypotheses with every finding linked to experimental evidence. **How is Plex different from literature-based AI tools?** Plex is an LLM for data and databases, not literature. While other tools search published text and generate summaries, Plex queries structured biological and chemical datasets directly. Every answer is traceable to specific data sources, with reproducible, evidence-scored outputs. **What is AutoPlex?** AutoPlex is Plex's autonomous AI research agent that plans and executes complex drug discovery research programs, interprets results, and provides full documentation of strategies and experimental validation data. **What data sources does Plex integrate?** Plex searches across 60+ databases including transcriptomics, proteomics, metabolomics, genomics, chemical structures, bioactivities, clinical studies, and scientific literature. Organizations can also securely integrate proprietary data. **Who uses Plex?** Plex is trusted by 40+ pharma and biotech discovery teams, including bench scientists, computational biologists, heads of translational science, and CSOs making portfolio decisions. **What therapeutic areas does Plex cover?** Plex has been used across oncology, CNS, immunology, rare disease, and inflammatory disease programs. The platform is indication-agnostic. **How long does a Plex evaluation take?** A typical program evaluation is delivered within weeks, not months — saving 4-6 weeks per program compared to conventional approaches.