cirrocumulus 1.1.17.post2

Single-cell visualization application

Cirrocumulus

Cirrocumulus is an interactive visualization tool for large-scale single-cell genomics data, with the following features:

• Run on a laptop, on-premise server, cloud VM, or Google App Engine

• View spatial transcriptomics data overlaid on an image

• Share the current visualization state in a URL

• Share datasets securely with collaborators

• Create dotplots, heatmaps, and violin plots to explore relationships between categorical variables and expression

• Explore complete differential expression results generated by Scanpy or Pegasus/Cumulus

• Interactively create and share “AND” or “OR” filters

• Collaboratively annotate cell types in real time

• Quickly load multiple features from predefined lists

• Explore multiple features and embeddings simultaneously

• Fast interactive exploration of 2 and 3-d embeddings of millions of cells, including zoom, pan, rotate (3-d), and lasso tools

• Save publication quality images

• Highly customizable - for example, set the color map, point size, or whether to use fog for 3-d embeddings to fade distant points

• Visualize datasets in h5ad, loom, Seurat, or STAR-Fusion formats

Quick Start

Install the package:

pip install cirrocumulus

Launch cirrocumulus via the command line:

cirro launch <path_to_dataset>

• Datasets can be provided in h5ad, loom, Seurat, or STAR-Fusion formats

• Launch accepts more than one dataset to enable quick dataset switching or to combine modalities (e.g gene fusions and expression) stored in separate files.

• Predefined marker lists can be provided in JSON format (see example) to quickly browse features of interest.

Example Data

• Download 3k PBMCs from a healthy donor data and launch:

  cirro launch pbmc3k.h5ad --markers markers.json

• Download human lymph node spatial data and launch:

  cirro launch V1_Human_Lymph_Node/V1_Human_Lymph_Node.h5ad --spatial V1_Human_Lymph_Node/spatial

Complete documentation available at https://cirrocumulus.readthedocs.io