Drug discovery is drowning in data. Literature databases contain millions of papers. Protein databases catalog hundreds of thousands of structures. Clinical trial registries track tens of thousands of studies. Yet the most promising therapeutic targets often hide in the gaps between these sources - overlooked because no single database captures the full picture.
TargetDiscovery Compass changes that. It integrates literature, omics, and structural data to map disease-driven pathway rewiring and surface tractable targets that traditional approaches miss.
The KRAS Challenge
In pancreatic ductal adenocarcinoma (PDAC), 90-92% of tumors harbor mutations in KRAS [1]. For decades, KRAS was considered "undruggable" - its smooth protein surface lacked binding pockets for small molecules. The emergence of covalent inhibitors like sotorasib and adagrasib has renewed enthusiasm, but resistance develops quickly and combination strategies are needed [2].
The problem isn't just hitting KRAS directly. It's understanding the entire network of pathway rewiring that KRAS mutations trigger - and finding the overlooked nodes where intervention might be most effective.
Integrating Thousands of Sources
TargetDiscovery Compass pulls evidence from across the research landscape:
Literature: PubMed, Europe PMC - millions of publications describing mechanisms, phenotypes, and experimental results
Protein Biology: UniProt - comprehensive protein function, interactions, and disease associations
Structures: PDB - 3D protein structures revealing druggable binding sites
Small Molecules: ChEMBL, PubChem - compound bioactivity data and chemical properties
Clinical Landscape: ClinicalTrials.gov - ongoing trials, approved drugs, and failed attempts
Intellectual Property: Lens.org, Google Patents - competitive landscape and freedom to operate
Each source provides a piece of the puzzle. TargetDiscovery Compass computes evidence strength and pathway interactions automatically across all of them.
From Data to Mechanism
The real value emerges when these sources are integrated into mechanistic causal graphs:
This network visualization shows KRAS at the center, with causal connections to downstream pathways: MAPK signaling (RAF, MEK, ERK), PI3K/AKT (mTOR, AKT, PI3K), Autophagy (LC3, ATG7, ULK1), Metabolism (GLUT1, PKM2, LDHA), DNA Damage (ATM, CHK2, p53), and Chromatin regulation (EZH2, BRD4, RRM2).
Edge weights represent causal strength computed from integrated evidence. Strong links (>0.7) indicate high-confidence mechanistic relationships. The highlighted target - identified by Compass as overlooked yet tractable - represents an opportunity that traditional single-source analysis would miss.
Surfacing Overlooked Targets
Why do promising targets get overlooked? Several reasons:
Literature fragmentation: Key findings may be scattered across papers in different subfields that rarely cite each other.
Database silos: A protein might have excellent structural data in PDB but limited functional annotation in UniProt.
Clinical blind spots: Targets validated in one disease context may be unexplored in related indications.
Patent thickets: Freedom-to-operate concerns may have discouraged development despite strong biological rationale.
TargetDiscovery Compass addresses these gaps by computing cross-source evidence scores. A target scores high when multiple independent data types converge: structural tractability, genetic association, pathway centrality, and limited competitive activity.
The ATM Example
Consider ATM kinase in KRAS-mutant PDAC. ATM sits in the DNA Damage pathway, responding to replication stress that KRAS-driven proliferation induces. Literature evidence supports synthetic lethality with KRAS mutations [3]. Structural data shows druggable binding sites. Yet ATM inhibitors remain underexplored in pancreatic cancer compared to other contexts.
TargetDiscovery Compass surfaces ATM by integrating:
- Literature co-occurrence with KRAS and PDAC
- Protein interaction networks showing pathway connectivity
- Clinical trial activity (limited in PDAC, active elsewhere)
- Patent landscape (composition of matter available)
- Structural druggability scores
No single source would prioritize ATM. The integrated evidence makes the opportunity clear.
Combination Strategy Insights
KRAS inhibitors alone face rapid resistance [4]. TargetDiscovery Compass helps identify rational combinations by mapping pathway dependencies:
KRAS + MEK: High pathway connectivity (0.90 edge weight) but overlapping mechanism may limit benefit.
KRAS + Autophagy: Moderate connectivity (0.60) but autophagy upregulation is a known resistance mechanism, suggesting synthetic lethality.
KRAS + Metabolism: Strong connectivity (0.75) to metabolic reprogramming - GLUT1 and PKM2 may offer orthogonal intervention points.
These insights emerge from the causal graph structure, not from any single publication or database.
From Weeks to Hours
Traditional target discovery involves months of literature review, manual database queries, and expert consultation. TargetDiscovery Compass compresses this timeline by automating evidence integration while preserving scientific rigor.
Researchers specify a disease context (e.g., KRAS-mutant PDAC), and Compass returns ranked target candidates with supporting evidence from all integrated sources. Each recommendation includes confidence scores, pathway context, and competitive landscape analysis.
Applications
Lead Target Identification: Surface overlooked targets with strong integrated evidence for new drug discovery programs.
Indication Expansion: Identify diseases where existing compounds might be repurposed based on pathway biology.
Combination Rationale: Map pathway dependencies to design mechanism-based combination strategies.
Competitive Intelligence: Understand where competitors are investing and where white space remains.
Drug discovery succeeds when insight connects disparate evidence. TargetDiscovery Compass makes those connections visible, surfacing the targets that matter most.
References
[1] Waters AM, Der CJ. "KRAS: The Critical Driver and Therapeutic Target for Pancreatic Cancer." Cold Spring Harb Perspect Med. 2018. https://www.frontiersin.org/journals/medicine/articles/10.3389/fmed.2024.1369136/full
[2] NCI. "Chemo Boosts KRAS Inhibitor Against Pancreatic Cancer." Cancer Currents. 2024. https://www.cancer.gov/news-events/cancer-currents-blog/2024/pancreatic-cancer-kras-inhibitors-chemotherapy
[3] Perkhofer L, et al. "ATM Deficiency Generating Genomic Instability Sensitizes Pancreatic Ductal Adenocarcinoma Cells to Therapy-Induced DNA Damage." Cancer Res. 2017.
[4] MSK. "Discovery Suggests Opportunity to Improve Effectiveness of KRAS Inhibitors Against Pancreatic Cancer." 2024. https://www.mskcc.org/news/msk-discovery-suggests-opportunity-to-improve-effectiveness-of-kras-inhibitors-against-pancreatic