Semantic Coherence: Reactome – Signal Evidence & AI Readability

[H3] Find Reactions, Proteins and Pathways

Introducing Reactome’s new Pathway Browser! Click the button to explore the new Pathway Browser Beta!Don’t forget to read our Release Notes, and share your feedback!

[H2]

Pathway Browser

Visualize and interact with Reactome biological pathways

[H2]

Analysis Tools

Merges pathway identifier mapping, over-representation, and expression analysis

[H2]

AI Chatbot

Meet the React-to-Me AI Chatbot! Designed to answer your questions about Reactome Pathways.

[H2]

ReactomeFIViz

Designed to find pathways and network patterns related to cancer and other types of diseases

[H2]

Documentation

Information to browse the database and use its principal tools for data analysis

[H3] Reactome Research Spotlight

[April 13, 2026] In their March 2025 Society of Toxicology paper “A workflow for human health hazard evaluation using transcriptomic data and Key Characteristics-based gene sets”, Tsai et al. present a systematic workflow for linking transcriptomic data to chemical hazard identification using Key Characteristics (KCs). Reactome and KEGG pathways were mapped to 34 umbrella KC terms collated from seven published hazard KC sets, generating 34 "KC gene sets" for each database for enrichment analysis. Tested against gene expression data for benzene, TCDD, sunitinib, and a negative control, the KEGG- and Reactome-derived KC gene sets successfully identified mechanistically relevant hazard terms - demonstrating a transparent, reusable approach for integrating transcriptomics into regulatory toxicology.
MORE

[H3] Why Reactome
Reactome is a free, open-source, curated and peer-reviewed pathway database. Our goal is to provide intuitive bioinformatics tools for the visualization, interpretation and analysis of pathway knowledge to support basic research, genome analysis, modeling, systems biology and education.

[IMG: European Bioinformatics Institute (EMBL-EBI)]

[IMG: NYU Langone Health]

[IMG: Oregon Health & Science University]

[IMG: Ontario Institute for Cancer Research]

The development of Reactome is supported by grants from the US National Institutes of Health (U24 HG012198) and the European Molecular Biology Laboratory.

[H3] Latest News

V96 Released

Reactome Releases Two New AI-Focused Preprints

New Publication in NAR 2026:

V95 Released

Reactome Pathway Browser: New Beta Release

V94 Released

V93 Released

Reactome Recognized with CoreTrustSeal Certification

[H3] Version 96 released on April 2, 2026

[H3]

[H2]
2,870
[H2]

Human Pathways

[H2]
16,338
[H2]

Reactions

[H2]
11,677
[H2]

Proteins

[H2]
2,198
[H2]

Small Molecules

[H2]
1,102
[H2]

Drugs

[H2]
42,784
[H2]

Literature References

[H3] Do you need help ?

[H2]

Use Reactome!

[H2]

For Developers

[H2]

Citing us

[H2]

Contact Us

[H3] API and Data access

[H2]

Content Service

[H2]

Analysis Service

[H2]

Icon Library

[H2]

Graph Database

© 2026
Reactome

This website requires cookies and the limited processing of your personal data in order to function. By using the site you are agreeing to this as outlined in our Privacy Notice. I agree, dismiss this banner

Cite Us!

Cite Us!

Warning! Unable to extract citation. Please try again later.

[IMG: Reactome logo]

[IMG: Reactome outage icon]

Ooops!
This is unexpected...
An error has occured and we are working to fix the problem.
We'll be up and running shortly.
Thank you for your patience.
If you need immediate assistance please contact
help@reactome.org.

Try again

Home

>

Docs

>

Userguide

>

Analysis Tools

UserguideDeveloper's ZoneIcon InfoData ModelCurator GuideRelease DocumentationComputationally inferred eventsFAQLinking to UsCiting us

[H4] Jump to section:

Analysis Data
Analysis Gene Expression
Species Comparison
Tissue Distribution

See our Youtube Video that introduces our Analysis Tools!
[H4] Analysis Data

Click on the ‘Browse Pathways’ button on the Homepage. In the next page, click the ‘Analysis’ button on the top right:

Alternatively, select the ‘Analyze Data’ button on the Homepage:
[IMG: analyze data button]

This opens a submission form, where you can select the analysis you want to perform, paste in or browse to a file containing your data, or use an example data set.
[IMG: analysis 3]

There are two sections to the submission form. The ‘Analyse your data’ section is selected by default to submit your data. Several different analyses can be performed, depending on the format of your data.
If your data is a single column of identifiers such as UniProt IDs, gene symbols or ChEBI IDs, they are mapped to pathways and over-representation and pathway-topology analyses are run. Over-representation analysis is a statistical (hypergeometric distribution) test that determines whether certain Reactome pathways are over-represented (enriched) in the submitted data. It answers the question ‘Does my list contain more proteins for pathway X than would be expected by chance?’ This test produces a probability score, which is corrected for false discovery rate using the Benjamani-Hochberg method.
Pathway topology analysis considers the connectivity between molecules that is represented by the pathway steps (which we refer to as reactions) in the pathway. It groups all the molecules represented in each reaction as a pathway ‘unit’. If any of these molecules are represented in your query set, this is considered a match to that reaction. This may give a better indication of the proportion of the pathway that matches your data, rather than the number of molecules that are common between your data and the pathway. It may also indicate that your data matches the start, end or a particular branch of a pathway process. This test does not have a probability score.
If your data has one or more additional columns of numbers it will be recognized as expression data and expression data overlay will be performed. Note that this data format should include a header row. The first column header should start with the # symbol. The numbers are used to produce a scaled coloured overlay over Reactome pathway diagrams, as a means to visualize relative expression levels. Note that the numeric values do not have to be expression data, for instance by using gene association scores the same analysis can be used to visualize genotyping results.

[H4] Identifier mapping

The submission process recognizes many types of identifiers. As part of the pre-analysis, they are mapped to Reactome molecules. The ideal identifiers to use are UniProt IDs for proteins, ChEBI IDs for small molecules, and either HGNC gene symbols or ENSEMBL IDs for DNA/RNA molecules, as these are our main external reference sources for proteins and small molecules. Many other identifiers are recognized and mapped to appropriate Reactome molecules. Accepted identifiers include HUGO gene symbols, GenBank/EMBL/DDBJ, RefPep, RefSeq, EntrezGene, MIM, InterPro, EnsEMBL protein, EnsEMBL gene, EnsEMBL transcript, and some Affymetrix and Agilent probe IDs. UniProt isoforms may be specified using the format P12345-2. If the -n suffix is omitted, this canonical form and all isoforms of it will be matched. Mixed identifier lists (different protein identifiers or protein/gene identifiers) may be used. Identifiers must be one per line. Protein-specific identifiers will typically map to protein entities, while gene-specific identifiers will map to the gene, transcript and derived proteins. If desired results can be filtered to show protein-specific or gene/transcript-specific results, details below.
Below is an example of the identifiers-only format:
Click the Continue button. A second options selection page appears:
[IMG: analysis 4]

Project to human is checked by default. With this option selected, all non-human identifiers in your query are converted by the analysis service to their human equivalents. In general, this maximizes the chances of a successful match to Reactome’s curated human pathways. However, if you want to use non-human identifiers and match these to our computationally-inferred non-human pathways, uncheck the box. You may also choose to uncheck this box if your query consists of a mixture of human and microbial identifiers and your goal is to find pathways that represent the processes of infection.
‘Include Interactors’ is unchecked by default. With this box unchecked, your query will consider only Reactome pathways. If you choose to check the box, your query will consider Reactome pathways that have been expanded by including all available protein-protein interactors from the IntAct database. This greatly increases the size of Reactome pathways, which maximizes the chances of matching your submitted identifiers to the expanded pathway, but will include interactors that have not undergone manual curation by Reactome and may include interactors that have no biological significance, or unexplained relevance. In practice it is preferable to query with ‘Include Interactors’ unchecked in the first instance, followed by a repeated query with ‘Include interactors’ selected, if a substantial proportion of the submitted identifiers do not match a Reactome pathway, to see if they can be identified as interactors.

[H4] Results for Identifier lists without associated numeric values

If you submit a single column of protein or small molecule identifiers they are mapped to pathways and over-representation and pathway-topology analyses are performed. The results will resemble the example below.
[IMG: analysis 6]

Analysis results are shown in the Analysis tab, within the Details Panel. All Reactome pathways are shown, in blocks of 20 pathways, ranked by the p-value obtained from over-representation analysis. If multiple pathways have the same p-value, they are ranked by the number of identifiers in the query that match the pathway. The number of molecules matched/total number of molecules and FDR values are added to the right side of pathway names in the Hierarchy Panel. The names of reactions that match at least one identifier in the query, representing positive pathway topology analysis hits, are boxed in orange.
In the Analysis tab, clicking on the name of a pathway will select it in the Hierarchy, which if necessary will expand hidden hierarchical levels to show the pathway, while the name becomes highlighted in dark blue.
By default, molecules of all types (protein, small molecules, genes, transcripts) are used for over-representation analysis, but it is possible to restrict the analysis results to a specific subtype by using a drop-down list located top-left of the results table. Selecting one of the subsets will display results that consider only the selected molecular subtype.

The columns in analysis details represent:

Pathway name: Click the name to open the pathway.
Entities found: the number of curated molecules of the type selected with Results Type that are common between the submitted data set and the pathway named in column 1. Click on this number to display the matched submitted identifiers and their mapping to Reactome molecules.
Entities total: The total number of curated molecules of the type selected with Results Type within the pathway named in column 1.
Interactors found (if this option was selected). The number of interactor molecules of the type selected with Results Type that are common between the submitted data set and the pathway named in column 1. Click on this number to display the matched submitted identifiers and their mapping to Reactome molecules.
Interactors total (if this option was selected): The total number of interactor molecules of the type selected with Results Type within the pathway named in column 1.
Entities ratio: Put simply, the proportion of Reactome pathway molecules represented by this pathway. Calculated as the ratio of entities from this pathway that are molecules of the type selected with Results Type Vs. all entities of the type selected with Results Type.
Entities pvalue: The result of the statistical test for over-representation, for molecules of the results type selected.
Entities FDR: False discovery rate. Corrected over-representation probability.
Reactions found: The number of reactions in the pathway that are represented by at least one molecule in the submitted data set, for the molecule type selected with Results Type.
Reactions Total: The number of reactions in the pathway that contain molecules of the type selected with Results Type.
Reactions ratio: Put simply, the proportion of Reactome reactions represented by this pathway. Calculated as the ratio of reactions from this pathway that contain molecules of the type selected with Results Type Vs. all Reactome reactions that contain molecules of the type selected with Results Type.
Species Name.

When an analysis has run, the Pathway Browser will display the Pathway Overview. All pathways that contain identifiers from your submitted list are highlighted, using a coloured scale to indicate the corrected probability (FDR). The colour scheme can be changed using the colour profiles tab (artists easel icon) on the pop-out Settings panel, found on the right-hand edge of the pathway panel.  Selecting coverage in the Overexpression panel will alter the Pathway Overview display to show additional event crosslinks corresponding to areas that are heavily covered vs regions that have lower coverage (as shown in the pValue view), and making it easier to visualize the pathways enriched within your dataset.
The highlighting of pathways in the Overview provides an at-a-glance representation of analysis results for all pathways. To see the details of a specific pathway, double-click the node representing the pathway in the overview or in the Pathway Hierarchy on the left. Alternatively, click it once to select it and use the Show All button (square with outward pointing triangles inside) in the top left corner of the Overview panel. The Overview can be navigated using the mouse scroll wheel to zoom in and out and click and drag to move it around. Alternatively, use the navigation buttons in the bottom right corner of the overview panel. At any level of the pathway, the diagram key can be found by clicking the compass symbol in the top right corner.
[IMG: analysis 8]

Enhanced high-level diagrams represent analysis results within the label for subpathways. The label background changes from blue to white, a yellow band is used to indicate the proportion of the pathway that is represented in the query dataset. A grey bar above the label indicates the number of pathway entities that are represented in the query dataset, the total number of entities in the pathway, and the FDR corrected probability score

In Pathway Diagrams, entities are re-coloured (yellow in the default colour scheme) if they were represented in the submitted data set. Complexes, Sets and Subpathway Icons are coloured to represent the proportion that is represented in the submitted identifier list. In the figure below, Insulin receptor is yellow indicating that is was in the submitted list. Insulin was not in the submitted list so it is not re-coloured. The complex of insulin:Insulin receptor is part re-coloured, part not, indicating that some molecules in the complex were represented in the submitted dataset while others were not.

If the Include Interactors option was checked, entities with interactors that were part of the submitted identifier list have a ribbon across the top right corner. In the example below, RASA1 is not a direct match with the submitted list of identifiers but has interactors that match the list. The interactors are displayed; the yellow overlay indicates the matching interactors. PTPN11 is a direct match and has interactors that match the list. A limited number of interactors are displayed to avoid crowding.

At the right side of the Analysis results details is a button indicating the number of submitted identifiers that were not successfully matched to molecules in Reactome. Click the button to produce a list.

[H4] Results for Identifier lists with associated numeric values (expression representation)

To run expression analysis, submit your data in a format that includes a first row of column headers. The header for column 1 must start with the # symbol. The first column must contain protein, compound or other suitable identifiers, such as probe IDs. All other columns must be numeric values, with no alphabetical characters. The analysis tool will interpret your data as expression data. The numeric values are used to colour objects in pathway diagrams. This view was created for microarray data, but any dataset that consists of a list of identifiers with associated numeric values can be used, e.g. quantitative proteomics, GWAS scores.

The tool is launched using the Analyse data button in the Pathway Browser header bar. Either paste your data into the submission form or browse to a saved file (or select an example file).
The figure below shows the correct data format. Each row must have an identifier in the first column (a header row is optional).
The submission process recognized many types of identifiers. As part of the pre-analysis, they are mapped to equivalent UniProt accessions or for small compounds to ChEBI IDs. These are the ideal identifiers to use with Reactome analysis tools. Other identifiers that are recognized and converted to UniProt equivalents include HUGO gene symbols, GenBank/EMBL/DDBJ, RefPep, RefSeq, EntrezGene, MIM and InterPro IDs, some Affymetrix and Agilent probe IDs, Ensembl protein, transcript and gene identifiers. Identifiers that contain only numbers such as those from OMIM and EntrezGene must be prefixed by the source database name and a colon e.g. MIM:602544, EntrezGene:55718. Mixed identifier lists (different protein identifiers or protein/gene identifiers) may be used. Identifiers must be one per line.
By default, all non-human identifiers are mapped to their human equivalents, unless the Project to the human checkbox is unselected.

After column 1, all other columns must contain numbers, representing expression or other values. Comma-separated values(CSV) and tab separated value (TSV) files can be used. When submitted, columns of numbers are considered as separate samples or experimental conditions. The

Take the most out of our tools and data analysis. All you need is our User Guide.

[H4]

For Users

Are you interested in programatically querying our data or integrating our Widgets ?

[H4]

For Developers

Have our data been useful in your research or experiment ? Please, remember to cite us!

[H4]

Citing us

The online User’s Guide begins with entry-level information that describes what our resource contains and how its pathways are organized. The guide leads users through the process of browsing Reactome, searching for particular biological and chemical molecules or pathways of interest, interpreting experimental datasets through the website and the more specialized ReactomeFIViz app.
[H3] Introduction to the Reactome Pathway Knowledgebase

Please visit our "Getting Started with Reactome" playlist to get an overview of Reactome!

[H3] Reactome Training Materials

All our training and outreach materials are available under a Creative Commons Attribution 4.0 Unported License. We have just two simple requests, please attribute Reactome, and let us know if you use our presentations, posters, or training resources. Here are some of the material that we make available to download:

Tutorial workshop slides. This document provides additional information about the Reactome website and suite of data analysis tools, including exercises to allow you to check your understanding. Answers are provided and were correct when written, but please note that answers may change over time as more pathways are added to Reactome.
Pathways and Networks Overview. These slides provide an overview of pathway and network resources for data visualisation, analysis, and integration.
Reactome FI Viz app. These slides provide an introduction to the Reactome Functional Interaction Network and Cytoscape app. These Reactome Questions and Answers provide a series of questions and answers to demonstrate the functionality of the Reactome FI Viz app.

Other online documentation includes complete descriptions of the Reactome data model and database schema, information for managers of external biological resources on how to link to specific types of Reactome pages, and information on how to cite the resource in publications. In addition, users can download specialized documentation that describes how to use the curator tool, a Reactome specific software tool for submitting data directly to the knowledgebase. One goal for Reactome has been the development of reusable software tools and data resources for managing and visualizing pathway information. This requires the creation of portable software and careful documentation. We believe in documenting our curatorial practices and policies as well, allowing other groups to understand and adopt our best practices.
As with extensions of Reactome and integration of new data and software, clear documentation of these methods are developed and, when possible, published in peer-reviewed journals. Documentation is actively maintained. We ask our broader base of users to join Reactome curators, technical writers, developers and educators beyond our primary institutions: OICR, OHSU, NYU, and EBI in developing, refining, and improving our documentation.

© 2026
Reactome

This website requires cookies and the limited processing of your personal data in order to function. By using the site you are agreeing to this as outlined in our Privacy Notice. I agree, dismiss this banner

Cite Us!

Cite Us!

Warning! Unable to extract citation. Please try again later.