Ragas

Ragas integrates and visualizes multi-level subclustering analysis of single-cell RNA-Seq data through a new data structure that connects and assembles analyses from different levels.

Key Features:

• Multi-level integration: A new data structure seamlessly connects and assembles miscellaneous single cell analyses from different levels of subclustering.
• Enhanced visualization: Several new or enhanced visualization functions are provided for streamlined analysis and presentation of subclustering results.
• Re-projection algorithm: An algorithm integrates nearest-neighbor graphs from multiple subclusters to maximize their separability on combined cell embeddings, significantly improving presentation of rare and homogeneous subpopulations.

Scientific Applications:

• Single-cell RNA-Seq subcluster analysis: Systematic integration of results from multiple subclusters enables accurate data quantification, visualization, and interpretation in downstream analysis.
• Rare subpopulation identification: The re-projection algorithm improves presentation of rare and homogeneous subpopulations that may be obscured in standard analyses.

Methodology:

Ragas implements a new data structure to connect and assemble analyses from different levels of subclustering, and a re-projection algorithm that integrates nearest-neighbor graphs from multiple subclusters to maximize their separability on combined cell embeddings.