MelOD (Melanoma Omics Dashboard) is a free, web-based interactive platform integrating preprocessed data from 16 melanoma studies, including eight bulk transcriptomics, six single-cell RNA-seq, and two proteomics datasets. MelOD provides user-friendly visualization and analysis tools—differential expression, dimensionality reduction, clustering, correlation, and survival analysis—without requiring local computational resources.
Several datasets include annotations for immunotherapy response, facilitating exploration of resistance and response signatures. Built on RShiny with optimized handling of large datasets, MelOD supports real-time hypothesis generation, cross-study validation, and community dataset contributions.
🔗 Live app: https://ruggleslab.shinyapps.io/MelOD/
If you use MelOD in your research, please cite our paper:
Sastourné-Haletou, P., Walker, A., Annuar, D., Subudhi, I., Karz, A., Berico, P., Angulo Salgado, P., Ibrahim, M., Osman, I., Schober, M., Hernando, E., & Ruggles, K. V. (2026). MelOD: The Melanoma Omics Dashboard for Multimodal Data Exploration. Pigment Cell & Melanoma Research, 39(4), e70101. https://doi.org/10.1111/pcmr.70101
📄 Paper: Pigment Cell & Melanoma Research
Melanoma poses a significant global health challenge due to its increasing incidence and high mortality rates. While immune-checkpoint blockade has transformed treatment in metastatic melanoma, approximately 40% of patients fail to respond or do not achieve durable benefit. Advances in high-throughput omics technologies have provided valuable insights into the molecular mechanisms underlying melanoma—but the complexity of these datasets and the lack of easily accessible melanoma-specific tools create significant barriers for researchers without computational expertise.
MelOD addresses these challenges with an accessible interface integrating core analytic workflows, providing a one-stop-shop for real-time data exploration, discovery, and collaboration.
- Free, web-based dashboard integrating data from 16 melanoma high-dimensional omics studies (over 1100 samples)
- Multi-modal support for bulk transcriptomics, single-cell RNA-seq, and proteomics visualization without local computational resources
- Immunotherapy response annotations enabling exploration of resistance and response signatures
- Modular architecture supporting community contributions and real-time cross-study validation
- Customizable volcano plots
- Heatmaps
- Gene–gene Pearson correlation matrices
- PCA plots
- Kaplan–Meier survival curves (when clinical metadata are available)
- Gene search across studies with interactive results panels and data tables
- Cell clustering based on low-dimensional embeddings (e.g., UMAP)
- Gene expression profiles and differential gene expression
- Cell type–specific marker genes, gene expression patterns, cell proportions
- Gene–gene co-expression plots
- Cell subsetting based on metadata or expression levels
- Download processed data, publication-ready figures (PNG, PDF), and analysis summaries
MelOD integrates preprocessed data from 16 melanoma-focused omics studies, encompassing over 1100 samples:
| Data Type | Number of Studies |
|---|---|
| Bulk transcriptomics | 8 |
| Single-cell RNA-seq | 6 |
| Global proteomics | 2 |
Each dataset underwent systematic preprocessing tailored to its data type:
- Bulk transcriptomics: Processed with the Seq-N-Slide pipeline, using STAR alignment, featureCounts, and DESeq2 for gene-level quantitation and differential expression analysis.
- Proteomics: Processed with MaxQuant for normalization and quality control.
- Single-cell RNA-seq: Processed with Seurat, including filtering of low-quality cells, normalization, and dimensionality reduction (UMAP, t-SNE).
MelOD is hosted on shinyapps.io for simultaneous multi-user access. It is built in RShiny using packages including shinyDashboard for the dashboard framework and plotly for interactive visualizations. Building on ShinyCell, the application splits Seurat objects into smaller components, storing gene expression matrices in a memory-efficient HDF5 format with metadata and dimensionality reduction coordinates in separate structures—enabling rapid interaction with large datasets without substantial local resources.
MelOD is accessible at https://ruggleslab.shinyapps.io/MelOD/, and the source code is available at https://github.com/ruggleslab/MelOD.
To foster collaboration and maximize MelOD's impact, we encourage researchers to submit additional omics datasets for consideration. Contribution requests can be made via the dedicated link on the homepage.
Supported by the NYULH Metastasis Research Network Center (NIH/NCI 1U54CA263001) and the NYU Melanoma SPORE grant (NCI P50 CA225450).
This project is licensed under the MIT License — see the LICENSE file for details.