RCA2

RCA2 enhances analysis of single-cell RNA sequencing (scRNA-seq) data by performing supervised clustering using reference transcriptomes to improve robustness to batch effects and technical variability.

Key Features:

• Supervised clustering: Uses known reference transcriptomes to guide clustering and distinguish biological variation from technical noise.
• Reference projection / batch-effect mitigation: Projects query cells onto reference transcriptomes to reduce batch effects across datasets.
• Graph-based clustering: Employs graph-based clustering to enable scalable analysis of large scRNA-seq datasets.
• Downstream analysis modules: Includes integrated downstream analysis modules commonly applied in scRNA-seq studies.
• Reference panels: Provides human and mouse reference panels and supports generation of custom reference panels.
• Cell type-specific quality control (QC): Implements cell type-specific QC measures to maintain data quality for heterogeneous tissues.

Scientific Applications:

• Human bone marrow profiling: Demonstrated to improve clustering accuracy and reduce technical biases in human bone marrow scRNA-seq datasets.
• PBMC profiling (healthy individuals): Applied to peripheral blood mononuclear cell (PBMC) datasets from healthy donors to refine cell-type identification.
• PBMC profiling (COVID-19 patients): Used on PBMC datasets from COVID-19 patients to enhance resolution of cellular responses associated with disease.
• Large-scale multi-batch scRNA-seq studies: Designed to provide consistent results across cohorts and experiments collected in multiple batches.

Methodology:

RCA2 projects cells onto reference transcriptomes (reference projection) and performs supervised, graph-based clustering to mitigate batch effects and scale to large scRNA-seq datasets.