scGENA

scGENA performs gene coexpression network analysis on single-cell RNA sequencing (scRNA-seq) data to identify coexpressed gene modules, characterize cellular heterogeneity, and link modules to disease-associated phenotypes.

Key Features:

• Integrated preprocessing pipeline: Includes data exploration, quality control, normalization, imputation, and dimensionality reduction for scRNA-seq datasets.
• Handling of scRNA-seq technical challenges: Addresses transcript sparsity, replication noise, and outlier cell populations to improve downstream inference.
• Gene coexpression network analysis: Constructs networks from gene-gene pairwise relationships to identify coexpressed gene modules that relate to phenotype.
• Clustering and rare cell detection: Clusters cells by expression and coexpression patterns to reveal rare and complex cell populations.
• Application to complex disease datasets: Demonstrated on a human diabetic pancreas scRNA-seq dataset (1,600 cells, 39,851 genes) to uncover disease-associated gene modules.

Scientific Applications:

• Cell Type Identification: Identifies distinct cellular populations within heterogeneous tissues by clustering coexpression patterns.
• Disease Mechanism Exploration: Reveals gene modules linked to conditions such as diabetes to support investigation of molecular underpinnings.
• Phenotypic Analysis: Facilitates exploration of phenotypic differences at single-cell resolution by associating coexpression modules with cellular behavior.

Methodology:

Performs data exploration and quality control followed by normalization, imputation, dimensionality reduction, construction of gene coexpression networks from pairwise gene relationships, module identification, and clustering of cells.