UHN Arthritis Program And Cytosolve Create A Multi-Scale MolecularSystems ArchitectureTo Map Human Knee Osteoarthritis From Tissues To Evidence-Linked Reactions.

Partner Description

UHN’s Arthritis Program within the Division of Orthopedics is a clinical-research group focused on musculoskeletal disease mechanisms and translational strategies, including knee osteoarthritis.

Challenge

Human knee osteoarthritis biology spans multiple anatomical compartments and cell types, producing a dense and fragmented mechanistic literature. Prior to this collaboration, researchers faced persistent barriers:

  • Scale and complexity: OA mechanisms involve cartilage, synovium, subchondral bone, meniscus, fat pad, osteophytes, immune cells, and pain pathways—difficult to synthesize into one coherent view.
  • Evidence fragmentation: Thousands of papers contain valuable but dispersed pathway findings, making comprehensive mechanism discovery slow and inconsistent.
  • Limited traceability: Even when pathway diagrams exist, linking a specific interaction back to its originating experimental paper is often non-trivial, hindering reproducibility and debate.

How CytoSolve® Helped

CytoSolve® built a multi-scale molecular systems architecture of human knee OA through a supervised bioinformatics process designed for quality, transparency, and usability.

Key components of the approach included:

  • Large-to-refined literature curation: PubMed served as the primary source database (search through June 2017). An initial yield of 20,231 papers was screened and refined to 5,243 studies included for architecture construction (PRISMA workflow).
  • Multi-layer “3D” traversal of biology: The architecture is accessed through an interface that progresses from.
    • Layer 1: knee anatomy and components
    • Layer 2: component-specific cell types (e.g., chondrocytes within cartilage)
    • Layer 3 to n: ensembles and individual molecular reactions
    • Final layer: clickable linkage to the specific paper supporting each reaction
  • Rigorous extraction and representation rules: Molecular interactions were extracted using explicit quality-control rules, including prioritizing data in figures/Results, focusing on human OA knee cells (not normal cells), using human data when multiple species were present, and avoiding reliance on referenced external studies.
  • Standardized graphical notation: The architecture encodes interaction types with consistent notation (e.g., direct receptor–ligand binding vs evidence of increased expression through unspecified mechanisms), improving interpretability across the full map.
  • Community feedback for currency: The system is designed to remain current by enabling the broader research community to submit feedback that can be reviewed, refereed, and incorporated.

Key Benefits Realized

  • Comprehensive, integrative view of knee OA systems biology spanning tissue → cell type → molecular reaction layers
  • Evidence-traceable interactions, allowing users to identify the exact study supporting a given molecular relationship
  • Quality-controlled mechanistic extraction, improving confidence in included interactions and minimizing propagation of unsupported claims
  • Actionable framework for research and translation, supporting investigation of disease mechanisms and identification of potential therapeutic targets
  • Reusable educational scaffold, enabling new learning tools grounded in curated, navigable OA biology

Outcome

The UHN–CytoSolve® collaboration produced a multi-scale molecular systems architecture of human knee osteoarthritis that converts a vast literature into a structured, explorable, and evidence-linked resource. By enabling traversal from knee anatomy to cell types to molecular reactions—with direct access to the supporting papers—and by supporting community-driven updates, the architecture provides a durable platform for mechanistic discovery, target identification, and education in knee OA.