FlowGrid

FlowGrid performs density-based clustering of large single-cell RNA-seq (scRNA-seq) datasets to identify cellular populations and characterize cellular heterogeneity.

Key Features:

• Density-Based Clustering Algorithm: Employs a fast density-based clustering algorithm originally developed for flow cytometry to identify clusters by regions of high cell density.
• Scalability and Speed: Processes datasets at scale, for example reducing runtimes for ~1,000,000 cells from around 1 hour to approximately 5 minutes.
• Automated Parameter Tuning: Implements an automated parameter tuning procedure to optimize clustering parameters.
• Integration with Scanpy: Integrates into the Scanpy workflow for use within Scanpy-based analyses.

Scientific Applications:

• Large-scale scRNA-seq clustering: Enables clustering analyses of scRNA-seq datasets containing millions of cells.
• Cell type discovery: Facilitates identification of novel cell types and cellular populations from scRNA-seq data.
• Cellular heterogeneity and biological process analysis: Supports studies of cellular heterogeneity and complex biological processes at single-cell resolution.

Methodology:

Uses a fast density-based clustering algorithm originally developed for flow cytometry, combined with automated parameter tuning, and provides an implementation that integrates with Scanpy.