Reactome

Reactome is a free, open-access database of biological pathways, built around a data model where the core unit is not the pathway as a whole but the individual reaction. Proteins, protein complexes, nucleic acids, and small molecules participate in a given reaction as inputs, outputs, catalysts, or regulators; chaining connected reactions together is what forms a pathway in Reactome’s model, rather than a pathway being drawn first and reactions filled in afterward. The project is a collaboration led by Lincoln Stein and maintained by teams at the Ontario Institute for Cancer Research (OICR), Oregon Health & Science University (OHSU), EMBL-EBI, and NYU Langone Medical Center.

How content gets into Reactome

Reactome content is authored by PhD-level biologists working with the project’s editorial staff, and every pathway passes a structured peer-review process before it is published into a release — the same editorial discipline a journal article goes through, applied to a pathway diagram instead of a manuscript. Each reaction step is backed by a cited experimental reference, so a researcher can trace not just what the pathway claims but the specific publication that supports each individual step within it. This is what distinguishes a curated resource like Reactome from a computationally inferred or crowd-sourced pathway collection: the editorial review is the actual gate, not just a stated aspiration.

Species coverage: human-curated, computationally projected elsewhere

Reactome’s primary curation effort is human biology. Equivalent pathways in other species are generated computationally, by projecting the curated human reactions onto orthologous genes identified in that species’ genome — not independently curated from scratch for every organism Reactome covers. This matters for how confidently a non-human pathway result should be treated: it inherits the reliability of the underlying orthology call as well as the human curation it was projected from, which is a genuinely different evidentiary basis than a directly curated human pathway.

Tools and access

Reactome’s Pathway Browser provides interactive visualization down to the individual reaction and molecule level, with expression-data overlay so a researcher can view their own experimental results directly on the pathway diagram. ReactomeGSA supports multi-omics gene set analysis, and ReactomeFIViz is a Cytoscape application for disease-related pathway and network discovery. Data exports support standard formats — BioPAX, SBML, and a Neo4j graph database dump among others — so a pathway can be pulled into external analysis pipelines rather than only viewed on Reactome’s own site. The database updates on a quarterly release cycle, which, as with any versioned resource, means a data management plan or methods section citing Reactome results should record the release used.

Where Reactome fits among named repositories

Reactome and KEGG both serve as curated pathway resources and are frequently run in parallel on the same gene list, but they differ in curation model: Reactome’s peer-reviewed, reaction-level editorial process centers on human biology with orthology-based projection outward, while KEGG’s pathway maps span a broader range of organisms and are centrally maintained rather than externally peer-reviewed per pathway. Neither substitutes fully for the other, and an enrichment analysis reporting results from only one without naming which was used, and its release version, is under-specifying a real methodological choice — the same accession-level specificity CASRAI’s own Data Management Plan guidance recommends for any named repository dependency.