Publication
Grace Potter, Jacob A Beierle, Camron Bryant, Sadhna Phanse, Carl White, Andrew Emili, Indranil Paul, OmNI: a modular open-source framework for interactive multi-omics data integration and visualization, NAR Genomics and Bioinformatics, Volume 8, Issue 1, March 2026, lqaf206, https://doi.org/10.1093/nargab/lqaf206
An interactive shiny application version of the original Omics Notebook, with increased functionalities.
Omics Notebook Interactive (OmNI) is an R-based, open-source, and modular framework engineered for streamlined multi-omics data integration and analysis across diverse data types, incorporating interactive visualizations at each processing step. OmNI performs differential expression analysis utilizing customizable linear models, accommodating various covariates and complex experimental designs. For cross-omic layer integration, OmNI employs a modified S-score statistic, ensuring sensitive detection of differential features. The framework also integrates network and metabolomics data, offering detailed insights into regulatory mechanisms through comprehensive enrichment analysis using multiple pathway databases. Outputs include interactive HTML reports, CSV/TSV files, and Cytoscape-compatible objects. OmNI is readily deployable in both local and high-performance computing environments, enabling scalable data processing.
An example dataset is included in the Data Upload tab so users can explore the workflow without uploading their own files.
Run locally:
# Launch the app from GitHub
shiny::runGitHub("OmNI", "gracerhpotter")# Download ZIP and run from folder
shiny::runApp("path/to/OmNI")Or access the deployed version:
👉 https://emili-laboratory.shinyapps.io/omni/
The navigation bar includes an About dropdown and six primary workflow tabs.
Contains introductory and reference information:
- Overview: Summary of OmNI’s capabilities and intended workflow.
- File Inputs: Expected input formats and data layout examples.
- Linear Modeling: Description of limma-based analysis steps.
- Enrichment: Overview of supported enrichment databases.
- Documentation: Links to R package references and methods.
Upload and explore your datasets or use the built-in Example Data.
Functionalities:
- Upload Data: Import
.csv,.tsv, or.txtexpression file(s) and an annotation file (.xlsx). - Missing Values: Explore missingness using VIM plots.
- Statistical Summary: Review descriptive statistics.
- Expression Matrices: Compare raw and normalized matrices.
- Normalization Plots: Visualize quality control via violin, QQ, RLE, and density plots.
There are a number of options for cutomization of data processing. Most options have preselections, and there are many tooltips providing information on the parameters. Once data is uploaded be sure to check that the Group selection (pulled from the provided annotation file) aligns with the desired analysis.
Perform exploratory analysis and assess dataset structure.
Tabs:
- Correlation Matrix: Pairwise sample correlation heatmap.
- MD Plot: Mean-difference distribution by contrast.
- UMAP: Low-dimensional visualization of sample relationships.
- Heatmap: Hierarchical clustering of samples/features.
- PCA: Principal component visualization of group separation.
Conduct differential analysis within a single omic layer.
Functionality:
- Limma Modeling: Apply linear modeling to detect significant features.
- Fit Summary: Interactive topTable of model statistics.
- Volcano Plot: Fold-change vs. significance visualization.
- Differential Heatmap: Cluster significant features.
- Enrichment: Perform functional enrichment using clusterProfiler.
- Supports GO, KEGG, Reactome, WikiPathways, MSigDB, and others.
- Results shown as dot plots, enrichment tables, gene set networks, and other visualizations from
clusterPofiler.
Integrate multiple omic datasets to identify shared biological patterns.
Features:
- S-Score Integration: Weighted combination of multi-omic significance.
- Summary Table: Lists integrated statistics and combined directionality.
- Volcano Plot: Display integrated differential results.
- Venn Diagram: Highlight shared and unique omic hits.
- Enrichment: Combined enrichment for integrated feature sets.
Build molecular networks and perform cluster-level enrichment.
Modules:
- PCSF Networks: Generate Steiner-forest-based functional subnetworks.
- PPI Network: Visualize protein–protein interactions.
- Influence Network: Highlight highly connected or influential nodes.
- Clustered Enrichment: Identify and annotate enriched subnetworks.
Compile and export all results.
📸 Screenshot placeholder:
- Click Use Example Data on the DATA tab.
- Proceed through Tabs 1–5 sequentially.
- Review each output interactively.
- Generate and download the full HTML report.
- Tabs are designed for sequential use but can be revisited at any time.
- Hover over plots for tooltips and data details.
- Large datasets may take several minutes; progress indicators show job status.
Dependencies for the application are installed and loaded via the loads.R file. The first time opening the application may take 10-15+ minutes if your system does not have many of the packages installed.
Additional information on the contents of each file, and the source of referenced databases for enrichment, PCSF, and S-Score is in the info folder in devInfo.txt. Additional reference scripts for the creation of some the files are also in that folder.
Example data and annotation files are present in the example folder, but can also be found in a public Google Drive folder.