ASURAT

ASURAT performs simultaneous unsupervised clustering and biological interpretation of single-cell RNA-seq data in terms of cell type, disease, biological process, and signaling pathway activity by transforming gene expression tables into sign-by-sample matrices using knowledge-based databases.

Key Features:

• Unsupervised Clustering: Performs unsupervised clustering of single-cell RNA-seq data without requiring prior cell type labels.
• Biological Interpretation: Simultaneously interprets clusters in terms of cell type, disease, biological process, and signaling pathway activity.
• Knowledge-Based Database Integration: Utilizes databases including Cell Ontology, Gene Ontology, and KEGG to inform analysis.
• Sign-by-Sample Matrix Generation: Transforms gene expression tables into original multivariate tables termed sign-by-sample matrices (SSMs).

Scientific Applications:

• Cell Type Identification: Identifies and characterizes cell types from single-cell RNA-seq data through knowledge-driven annotation.
• Disease State Analysis: Interprets single-cell data in the context of disease-related signatures and states.
• Biological Process Characterization: Associates cell clusters with specific biological processes using Gene Ontology and other functional databases.
• Signaling Pathway Activity Inference: Infers signaling pathway activity levels across single-cell populations using KEGG and related pathway databases.

Methodology:

ASURAT inputs single-cell RNA-seq data and knowledge-based databases (Cell Ontology, Gene Ontology, KEGG, etc.) and transforms gene expression tables into sign-by-sample matrices (SSMs).