muscat

muscat detects subpopulation-specific state transitions from multi-sample multi-condition single-cell transcriptomics data.

Key Features:

• Differential State Analysis: Provides methods for DS analysis in multi-sample, multi-group, multi-(cell-)subpopulation scRNA-seq data.
• Cell-Level Mixed Models: Implements cell-level mixed models for cross-condition differential state analysis.
• Pseudobulk Methods: Includes methods based on aggregated "pseudobulk" data for differential state analysis.
• Simulation Platform: Provides a flexible simulation platform that mimics both single and multi-sample scRNA-seq data.
• Bulk-Based Hypothesis Weighing: Implements bbhw to increase power and robustness of single-cell differential state analysis using larger bulk RNAseq data.
• Grouped Benjamini-Hochberg Procedure: Applies grouped Benjamini-Hochberg procedure on bins based on proportion-adjusted significance for optimal performance with bbhw.
• Visualization Tools: Provides visualization tools for differential state analysis results.

Scientific Applications:

• Subpopulation-Specific Response Detection: Uncovers subpopulation-specific responses to treatments such as lipopolysaccharide in mouse cortex cells.
• Multi-Condition Differential State Analysis: Enables sample-level inferences from replicated multi-condition scRNA-seq datasets.
• Cross-Condition Differential Expression: Compares expression levels across sets of cells for identifying subpopulation markers and state transitions.
• Power and Robustness Enhancement: Increases power and reproducibility of single-cell differential state analysis by leveraging available bulk RNAseq data.

Methodology:

muscat implements cell-level mixed models and methods based on aggregated "pseudobulk" data. For bulk-based hypothesis weighing, methods include grouped Benjamini-Hochberg procedure on bins based on proportion-adjusted significance.