cellucid - Interactive Single-Cell Data Visualization

GPU-powered atlas viewer for single-cell data with fast rendering, real-time filtering, and collaborative annotation.

Key Features

• Real-time rendering of millions of cells with adaptive level of detail (LOD)
• Gene expression overlays with efficient sparse matrix handling
• Categorical and continuous cell metadata coloring, filtering, and selection
• Multi-dimensional support (1D timelines, 2D, 3D embeddings)
• Animated vector field overlay (velocity / drift) with GPU particle flow
• Community annotation voting with optional GitHub sync
• Connectivity edge visualization (KNN graph)
• Publication export: SVG (vector) + PNG (high-DPI)

Data Sources

Load prepared exports, AnnData .h5ad files, or .zarr stores directly in the browser, from a Python server, or from a Jupyter notebook.

Use Cases

Explore scRNA-seq atlases, query marker genes, compare embeddings, visualize trajectories, and build consensus community annotations.