CPPLS_MLP

CPPLS_MLP constructs cell-cell communication networks and identifies highly variable genes (HVGs) associated with intercellular communication from scRNA-seq and spatial transcriptomics data.

Key Features:

• Integration of scRNA-seq and spatial transcriptomics: Combines scRNA-seq and spatial transcriptomics datasets to capture spatial and single-cell expression patterns relevant to cell-cell communication.
• Identification of Highly Variable Genes (HVGs): Identifies HVGs that are closely linked to intercellular communication and cell-type-specific expression variability.
• Multiple Input Multiple Output (MIMO) analysis: Applies a MIMO framework to assess how neighboring cell types and intercellular signals affect differential expression of HVGs.
• Network construction via ligand-receptor interactions: Constructs cell-cell communication networks by analyzing ligand-receptor interactions at multiple spatial and cellular scales.
• Performance versus CCPLS: Demonstrates enhanced accuracy in network construction and identification of cell-type-specific HVGs compared with CCPLS.

Scientific Applications:

• Immunology: Maps intercellular signaling and HVGs to investigate immune cell interactions and immune processes.
• Developmental biology: Dissects cell–cell communication and HVG dynamics relevant to developmental processes and organismal homeostasis.

Methodology:

Uses scRNA-seq and spatial transcriptomics for data acquisition; constructs cell-cell communication networks by analyzing ligand-receptor interactions at multiple scales; identifies HVGs associated with intercellular communication using algorithmic approaches; and applies a Multiple Input Multiple Output (MIMO) framework to analyze effects of neighboring cell types on HVG differential expression.