Independent component analysis based gene co-expression network inference (ICAnet) to decipher functional modules for...

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Single-cell RNA sequencing (scRNA-seq) generates large datasets for biological discovery.
• Current gene co-expression network methods struggle with rare cell types and batch effects.
• Existing methods often overlook cell-specific co-expression patterns.

Purpose of the Study:

• To develop a novel bioinformatics method for scRNA-seq data analysis.
• To improve cell clustering and rare cell-type discovery.
• To enhance batch integration and biological interpretation of scRNA-seq data.

Main Methods:

• Developed Independent Component Analysis-based Gene Co-expression Network Inference (ICAnet).
• Decomposed scRNA-seq data into independent gene expression components.
• Inferred co-expression modules from these components.

Main Results:

• ICAnet demonstrated improved cell clustering and rare cell-type discovery.
• The method showed efficient performance for batch integration across diverse scRNA-seq datasets.
• ICAnet is robust to variations in library strategy, sequencing depth, and cell number.
• Identified potential diagnostic markers for acute myeloid leukemia from scRNA-seq data.

Conclusions:

• ICAnet is a robust and effective tool for scRNA-seq data analysis.
• The method enhances the discovery of rare cell types and biological insights.
• ICAnet offers improved cell clustering and batch correction capabilities.
• ICAnet has potential applications in disease diagnostics, such as acute myeloid leukemia.