scBubbletree

scBubbletree visualizes scRNA-seq data by identifying clusters of cells with similar transcriptomes and representing them as quantitative "bubbles" at dendrogram tips, preserving local and global cell distances and cell density distributions across samples.

Key Features:

• Overplotting resolution: The method addresses overplotting problems common in standard two-dimensional visualization approaches for high-dimensional scRNA-seq data.
• Quantitative cluster visualization: Clusters are visualized as bubbles at the tips of dendrograms (bubble trees) with quantitative summaries of cluster properties and relationships.
• Multiomic data integration: The method enables visualization of diverse cell attributes from multiomic single-cell experiments.
• Scalability: The method has been demonstrated on large scRNA-seq datasets including one with over 1.2 million cells.
• Stacked information display: Bubble trees are stacked with further cluster-associated information in a visually accessible format for quantitative assessment.

Scientific Applications:

• scRNA-seq data exploration: Quantitative visual exploration of single-cell RNA sequencing data for analysis of cell relationships and biological interpretation.
• Cell cluster relationship analysis: Examination of relationships between transcriptome-similar cell clusters in high-dimensional datasets.
• Cross-sample cell density comparison: Preservation and comparison of cell density distributions across multiple samples.

Methodology:

The method identifies clusters of cells of similar transcriptomes and visualizes such clusters as bubbles at the tips of dendrograms (bubble trees), with quantitative summaries of cluster properties and relationships.