CINS

CINS infers differential cell type interactions from single-cell RNA sequencing (scRNA-Seq) datasets to identify condition-specific interaction networks and their ligand mediators.

Key Features:

• Bayesian network analysis: Uses Bayesian network analysis to infer interaction structure among cell types across conditions.
• Regression-based modeling: Applies regression-based modeling to quantify differential interactions and associate them with molecular mediators.
• Integrated approach: Combines Bayesian network analysis with regression-based modeling to jointly identify differential interactions and the ligands mediating them.
• Ligand and protein identification: Pinpoints proteins, particularly ligands, that mediate inferred cell type interactions.
• scRNA-Seq input: Operates on single-cell RNA sequencing (scRNA-Seq) data as the primary input.
• Detects interaction changes beyond proportions/expression: Identifies changes in cellular communication not captured solely by differences in cell type proportions or gene expression levels.
• Empirical performance: Demonstrated improved performance relative to existing methods in the datasets reported in the original study.
• Validation on biological datasets: Validated inferred networks using additional scRNA-Seq experiments in aging mouse models.

Scientific Applications:

• Disease case-control comparisons: Identifies differential cell type interactions and mediating ligands between case and control conditions.
• Aging studies in mouse models: Detects condition-specific interaction networks and ligand mediators in aging mouse scRNA-Seq datasets.
• Mechanistic studies of cellular communication: Maps interaction networks and candidate ligand mediators to support investigations of disease mechanisms and condition-specific cellular processes.

Methodology:

Combines Bayesian network analysis with regression-based modeling applied to scRNA-Seq data to infer differential cell type interactions and to associate interactions with ligand proteins.