scTensor

scTensor identifies many-to-many cell–cell interactions from single-cell RNA-sequencing data by extracting triadic ligand-expression, receptor-expression, and ligand–receptor (L-R) pair relationships using non-negative Tucker decomposition.

Key Features:

• Non-Negative Tucker Decomposition (NTD2): scTensor applies NTD2 to decompose high-dimensional single-cell RNA-sequencing data into interpretable components.
• Triadic Relationship Extraction: The method extracts triadic relationships represented as hypergraphs that capture ligand-expression, receptor-expression, and related L-R pairs.
• Detection of Many-to-Many CCIs: scTensor identifies many-to-many cell–cell interactions that extend beyond pairwise L-R gene co-expression analyses.
• Single-cell RNA-sequencing Input: The analytical framework operates directly on single-cell RNA-sequencing data.

Scientific Applications:

• Immunology: Analysis of complex intercellular communication networks among immune cell types from scRNA-seq data.
• Developmental Biology: Characterization of multi-cellular interaction patterns during tissue development using scRNA-seq.
• Cancer Research: Identification of intricate cell–cell interaction networks in tumor microenvironments from scRNA-seq.

Methodology:

Input single-cell RNA-sequencing data; apply Non-Negative Tucker Decomposition (NTD2) to decompose the data; construct hypergraphs representing triadic relationships (ligand-expression, receptor-expression, and L-R pairs).