Ragas: integration and enhanced visualization for single cell subcluster analysis.

Area of Science:

• Computational biology
• Bioinformatics

Background:

• Subcluster analysis enhances clustering and characterization of single-cell RNA sequencing (scRNA-Seq) data.
• Existing tools lack systematic integration of multi-level subclustering results, hindering downstream analysis.

Purpose of the Study:

• To develop an R package, Ragas, for integrating and streamlining multi-level subclustering analyses in scRNA-Seq data.
• To address limitations in data quantification, visualization, and interpretation arising from disparate subclustering results.

Main Methods:

• Implementation of a novel data structure to connect and assemble analyses from various subclustering levels.
• Development of enhanced visualization functions for integrated subclustering data.
• Introduction of a re-projection algorithm to integrate nearest-neighbor graphs from multiple subclusters for improved cell embedding separability.

Main Results:

• Ragas provides a unified framework for analyzing and visualizing integrated multi-level subclustering results.
• The re-projection algorithm effectively maximizes the separability of subpopulations in combined cell embeddings.
• Significant improvement in the presentation and identification of rare and homogeneous cell subpopulations.

Conclusions:

• Ragas offers a robust solution for overcoming challenges in analyzing complex scRNA-Seq data with multiple subclusters.
• The package facilitates more accurate quantification, enhanced visualization, and clearer interpretation of single-cell data.
• Ragas improves the discovery and characterization of cellular heterogeneity in scRNA-Seq studies.