SingleCellSignalR

SingleCellSignalR infers ligand-receptor interactions underlying cellular networks from single-cell transcriptomics or proteomics data.

Key Features:

• Curated Ligand-Receptor Database: The package includes a curated ligand-receptor database for inference of intercellular interactions.
• Regularized LR-Score: A novel regularized score outperforms other scoring schemes while controlling false positives and enables assessment of confidence in predicted ligand-receptor interactions.
• Dual Inference Framework: SingleCellSignalR v2 provides both the original SingleCellSignalR LR-score and the statistical model from the BulkSignalR package.
• Cross-Species Compatibility: Version 2 enables users to work with any species, expanding beyond the Mus musculus and Homo sapiens limitation of version 1.
• Network Visualization: The package provides a unique network view integrating all intercellular interactions and a function relating receptors to expressed intracellular pathways.
• Multiple Output Formats: Analysis results are provided in a variety of tabular and graphical formats.

Scientific Applications:

• Intercellular Network Inference: SingleCellSignalR infers ligand-receptor interactions that underlie cellular communication networks from single-cell expression data.
• Tissue Layer Communication Analysis: The tool can discover oriented communication structures, as demonstrated on mouse epidermis data revealing communication from external to basal layers.
• Intracellular Pathway Analysis: The package relates receptors to expressed intracellular pathways to understand downstream signaling effects.

Methodology:

SingleCellSignalR v2 relies on a statistical model introduced in the BulkSignalR package and the original SingleCellSignalR LR-score. The v2 implementation wraps BulkSignalR fundamental classes to enable cross-species compatibility.