Welcome to The Fusion Immunome Atlas

ImmunoFusion is the largest database for researching fusion genes in cancer, integrating genomic data from over 10,000 pan-cancer patients with immune-oncology profiling across immunotherapy-treated and untreated cohorts.

Database Overview
Quick Navigation
Fusion Data
Explore Fusion Gene List
Cohort Data
Explore Cohort Details
Distribution
Frequency, Location, Composition
Comparison
TME, Signatures, Response
Association
Correlation, Fisher, Interactions
Risk Analysis
KM Curves, Cox Regression
Landscape
Oncoprint, Cohort View
About the Platform
App Data

The platform integrates comprehensive genomic data from over 10,000 pan-cancer patients encompassing projects such as The Cancer Genome Atlas (TCGA), Therapeutically Applicable Research to Generate Effective Treatments (TARGET), and Clinical Proteomic Tumor Analysis Consortium (CPTAC). Additionally, it includes data from more than 4,000 cancer samples from patients with 10 distinct cancer types who have been treated with Immune Checkpoint Inhibitors (ICIs).

App Updates

2025-08-15: Release version v1.0.

2025-06-10: Database updated for confident fusions.

2025-06-05: First draft on bioRxiv.
App Development

Crafted with R Shiny, bslib, and shinyWidgets. We are committed to delivering a flawless user experience.

Disclaimer

ImmunoFusion is for research purposes only. It does not use cookies or collect personal information. Clinical use is at your own risk.
Contacts
Cohort Characteristics

IO = Immune Checkpoint Inhibitor treated cohorts. See the Cohort page for per-cohort details and the fusion landscape.

Gene Freq
Gene Pairs Associated
Gene Distribution by Cohort
Gene Records
CSV
You can use the search boxes above or under each column header for quicker filtering. You can use mouse drag to adjust the column width. If the table width is not fully displayed, you can use TouchPad (arrow keys etc.) for quick scrolling.
Cohort Overview
Click any row to open the cohort detail modal with summary statistics and a fusion landscape.

Explore fusion data distribution by Frequency, Localization, and Composition.

  • Frequency: fusion frequency distribution across cohorts.
  • Localization: genome-wide breakpoint mapping (circos + lollipop).
  • Composition: fusion type, cancer type, and partner gene breakdowns.

Active global filters (cohorts, fusion type, Score, tools, genes, fusions) restrict which fusions are included. Adjust filters to narrow or broaden the fusion set visualized in each distribution.

Analysis Controls
Results
Analysis Controls
Results
Analysis Controls
Target Selection
Composition Dimensions
Group small categories as "Other"

Multi-Dimensional

Cross-tabulation heatmap
Sankey flow diagram
Composition Results

Compare tumor microenvironment (TME) estimates or IOBR immune signatures between Fusion+ and Fusion- samples within a single cohort.

  • TME: compare cell-type abundance by fusion status. Select specific gene(s)/pair(s) to define Fusion+, or leave empty to use all detected fusions (any fusion = Fusion+).
  • IOBR Signature: compare immune signature scores by fusion status. Same gene/pair selection logic as TME.

Active global filters (fusion type, Score, tools, genes, fusions) define Fusion+ status: a sample is Fusion+ only if it carries at least one fusion passing ALL active filters. Adjust filters to refine grouping.

All results are computed on demand when you click Run.

Analysis Controls
TME by Fusion Status
Analysis Controls
Signature by Fusion Status

Explore associations between gene fusions and molecular features, clinical response, or pairwise gene-gene relationships.

  • Correlation: fusion count vs TME/signature feature.
  • Response: fusion status vs clinical response (binary + RECIST).
  • Gene Interactions: co-occurrence / mutual exclusivity heatmap.

Active global filters (cohorts, fusion type, Score, tools, genes, fusions) define what constitutes a Fusion+ event --- a sample is Fusion+ only if it carries at least one fusion passing ALL active filters. Adjust filters to refine the Fusion+ vs Fusion- grouping.

All results are computed on demand when you click Run.

Analysis Controls
log10(x+1) fusion count
Feature vs Fusion Count
Analysis Controls
Response by Fusion Status
Analysis Controls
Gene Interaction Results

Assess the prognostic value of gene fusions using Kaplan-Meier survival analysis.

  • Gene-Specific Survival: KM curve for a selected gene/pair within one cohort.

Active global filters (fusion type, Score, tools) define Fusion+ status: a sample is Fusion+ only if it carries at least one fusion passing ALL active filters. Adjust filters to refine the Fusion+ vs Fusion- grouping.

All results are computed on demand when you click Run.

Gene-Specific Survival
Analysis Controls
Gene-Specific Survival

Assess the prognostic value of gene fusions using Cox proportional hazards regression.

  • Cox Forest Plot: multivariable Cox regression for selected gene(s)/pair(s) with optional clinical covariates (Sex, Age). Hazard ratios and 95% CI are plotted.

Active global filters (fusion type, Score, tools) define Fusion+ status for each gene: a sample is Fusion+ for a given gene only if the fusion passes ALL active filters. Adjust filters to refine the grouping.

All results are computed on demand when you click Run.

Cox Forest Plot
Analysis Controls
Cox Regression Forest Plot

Explore the fusion landscape within a single cohort as an interactive OncoPrint. Each column represents a sample; each row represents a gene or gene pair. Colored rectangles within cells indicate specific fusion types detected.

  • Alteration types: CodingFusion (blue), NoHeadGene (amber), SameGene (green), TruncatedCoding (red), TruncatedNoncoding (purple).
  • Hover over cells for details. Click a cell to see its fusion records.
  • Clinical annotations: select covariates to add sample-level annotation bars above the heatmap.

Active global filters restrict which fusions are included.

Analysis Controls
Show negative samples
Fusion Landscape
Original heatmap
Selected sub-heatmap
Output