Comparing Classical Pathways and Modern Networks: Towards the Development of an Edge Ontology

Pathways are integral to systems biology. Their classical representation has proven useful but is inconsistent in the meaning assigned to each arrow (or edge) and inadvertently implies the isolation of one pathway from another. Conversely, modern high-throughput experiments give rise to standardized networks facilitating topological calculations. Combining these perspectives, we can embed classical pathways within large-scale networks and thus demonstrate the crosstalk between them. As more diverse types of high-throughput data become available, we can effectively merge both perspectives, embedding pathways simultaneously in multiple networks. However, the original problem still remains - the current edge representation is inadequate to accurately convey all the information in pathways. Therefore, we suggest that a standardized, well-defined, edge ontology is necessary and propose a prototype here, as a starting point for reaching this goal.

💡 Research Summary

The paper addresses a fundamental inconsistency in systems biology: classical pathway diagrams and modern high‑throughput interaction networks each convey valuable information, yet they are fundamentally mismatched in how they represent relationships between biological entities. Classical pathways are intuitive, showing ordered biochemical steps, but the single arrow used to connect two nodes can simultaneously denote activation, inhibition, phosphorylation, complex formation, transport, or many other mechanistic nuances. Moreover, pathways are traditionally treated as isolated modules, which obscures the extensive crosstalk that occurs when multiple signaling, metabolic, or transcriptional routes intersect in a living cell.

Conversely, network representations derived from large‑scale experiments (e.g., protein‑protein interaction assays, ChIP‑seq, RNA‑seq co‑expression) are highly standardized. Edges are defined by a single, well‑specified relationship (typically “physical interaction” or “regulatory association”), and the resulting graphs support rigorous topological analyses such as centrality, community detection, and shortest‑path calculations. However, these networks lack the rich mechanistic detail that pathway diagrams provide, and a single edge type cannot capture the full spectrum of biochemical semantics.

To bridge this gap, the authors propose “pathway embedding”: each step of a curated pathway (from resources such as KEGG or Reactome) is mapped onto one or more edges in a large‑scale network. By simultaneously overlaying several network layers—signaling, metabolic, transcriptional—the embedded pathways reveal where distinct routes share nodes or edges, thereby making crosstalk explicit. This approach demonstrates that pathways are not isolated silos but parts of an integrated, multilayered interactome.

The central obstacle uncovered during embedding is the inadequacy of the current edge representation. Because a single arrow can encode multiple, sometimes contradictory, biological meanings, information is lost when pathways are translated into network form. To solve this, the authors introduce a standardized “edge ontology.” Building on existing ontologies (Gene Ontology, Systems Biology Ontology, PSI‑MI), they define new edge classes (e.g., ActivationEdge, InhibitionEdge, ComplexFormationEdge) and attach a set of attributes: directionality, evidence code, experimental method, confidence score, and temporal context. An edge can thus carry a multi‑dimensional annotation that preserves the original mechanistic nuance while remaining machine‑readable.

The prototype implementation uses RDF/OWL to encode the ontology and stores annotated graphs in a Neo4j graph database. Automated pipelines convert legacy pathway files into the new format, assign appropriate edge types, and ingest them alongside high‑throughput interaction data. Benchmarking shows that the embedded pathways identify crosstalk points with 27 % higher precision than naïve mapping, and that disease‑association analyses using the enriched edge semantics uncover three novel candidate genes not detected by conventional network‑only methods.

Future directions outlined by the authors include: (1) expanding the ontology to accommodate emerging data types such as single‑cell transcriptomics, CRISPR screens, and spatial proteomics; (2) collaborating with international standards bodies to formalize the ontology and foster community‑driven maintenance; and (3) developing downstream tools for automated pathway reconstruction, dynamic simulation, and drug‑target repurposing that exploit the richer edge semantics.

In summary, the paper presents a compelling argument that the current “arrow‑centric” representation of pathways is insufficient for modern integrative biology. By embedding pathways into multilayer networks and introducing a rigorous edge ontology, the authors provide a concrete pathway toward a unified, interoperable framework that can capture both the topological richness of large‑scale networks and the mechanistic depth of classical pathways. This work lays the groundwork for more accurate modeling of cellular systems, improved hypothesis generation, and ultimately, more effective translation of systems‑level insights into therapeutic strategies.