Cascade

Unleash the power of your single-cell data

Cascade is an interactive dashboard that transforms complex single-cell RNA-Seq and spatial transcriptomics data (Visium and Xenium) into beautiful, insightful visualizations. Designed for both computational and experimental biologists, Cascade makes exploring your data intuitive and exciting.

Key Features

• Universal Compatibility: Works seamlessly with Seurat (.Rds) and scanpy (.h5ad) objects
• Lightning-Fast Visualizations: Interactive plots powered by plotly that respond in real-time
  • Violin plots - Compare expression across clusters
  • Dot plots - Visualize expression prevalence and intensity
  • Feature plots - Map gene expression onto UMAP or spatial coordinates
  • Co-expression plots - Discover genes with similar expression patterns
  • Scatter plots - Explore relationships between genes
  • Line plots - Track expression trends across conditions
• Spatial Transcriptomics: Visualize gene expression in tissue context with Visium and Xenium support
• Interactive Cell Selection: Lasso select cells of interest directly on UMAP/spatial plots and export barcodes for downstream analysis
• Smart Filtering: Filter cells based on metadata and gene expression with an intuitive interface
• Cluster Relationships: Visualize hierarchical relationships between cell clusters with beautiful tree plots
• Gene Tracking: Keep your genes of interest at your fingertips with the "Gene scratchpad"
• Powerful Marker Analysis: Interactive marker tables with advanced filtering capabilities
• Flexible Deployment: Run locally for personal analysis or on a server to share with collaborators
• User Management: Optional authentication system for controlled access in multi-user environments

Installation

conda (recommended)

The easiest way to get started with Cascade is through conda, which handles all dependencies automatically:

# Create environment outside the cascadeR directory
cd .. &&  env create -p env --file cascadeR/requirements-pinned.yaml
conda activate ./env

Then install the package using one of these methods:

Option 1: Using remotes::install_github

Note that here we use upgrade='never' to leave the conda installed package versions unchanged.

remotes::install_github('NICHD-BSPC/cascadeR', upgrade='never')

Option 2: Using R CMD build and install.packages:

R CMD build cascadeR/
Rscript -e "install.packages('cascadeR_1.0.tar.gz', repos=NULL)"

remotes

Alternatively, install directly with remotes:

install.packages('remotes')
setRepositories(ind=c(1,2,3,4,5))  # Get both CRAN and Bioconductor packages
remotes::install_github('NICHD-BSPC/cascadeR')

Getting Started

Data Organization

Organize your data in a directory structure that Cascade can easily navigate:

/cascade/data/
  ├─ project1
  │  ├─ seurat
  │  │  ├─ clustered.Rds
  │  │  └─ allmarkers.tsv
  │  │
  │  └─ scanpy
  │     ├─ clustered.h5ad
  │     └─ allmarkers.tsv
  │
  └─ project2
     └─ seurat5
        └─ clustered.Rds

Launch Cascade

First, load the library and install required Python dependencies:

library(cascadeR)
install_cascade()  # Installs plotly and kaleido for interactive plots
run_cascade()      # Launch the app!

The first time you run Cascade, you'll be prompted to choose a data directory. Point it to your data location (e.g., /cascade/data), and you're ready to explore!

Remote Access

Running Cascade on a remote server? No problem:

run_cascade(options=list(port=12345, launch.browser=FALSE))

Then access the app at http://127.0.0.1:12345 through your SSH tunnel.

Server Mode with Authentication

For multi-user environments, Cascade supports authentication:

# Create user database
credentials <- data.frame(
  user = c('shinymanager'),
  password = c('12345'),
  admin = c(TRUE),
  stringsAsFactors = FALSE
)

# Initialize the database
shinymanager::create_db(
  credentials_data = credentials,
  sqlite_path = 'credentials.sqlite',
  passphrase = 'admin_passphrase'
)

# Run with authentication
run_cascade(credentials='credentials.sqlite', passphrase='admin_passphrase')

Exploring Your Data

Once your data is loaded, Cascade offers multiple ways to explore:

• Summary Tab: Get a quick overview of your dataset
• Cell Embeddings: Visualize cells in UMAP or spatial context
• Metadata Viewer: Explore cell metadata and cluster characteristics
• Cluster Tree: Understand relationships between cell clusters
• Cell Markers: Identify and filter marker genes for each cluster
• Marker Plots: Create beautiful visualizations of gene expression
• Settings: Configure data directories and user access

Documentation

Each module includes comprehensive help documentation accessible through the help buttons throughout the interface.

Contributing

We welcome contributions to Cascade! Please feel free to submit issues or pull requests to the GitHub repository.

License

Cascade is available under the MIT license.