scTenifoldNet

scTenifoldNet constructs and compares transcriptome-wide single-cell gene regulatory networks (scGRNs) from single-cell RNA sequencing (scRNA-seq) data to identify differentially regulated genes and shifts in gene expression programs.

Key Features:

• Machine Learning Integration: Employs principal component regression, low-rank tensor approximation, and manifold alignment for construction and comparison of scGRNs.
• Data Handling: Tailored to analyze single-cell RNA sequencing (scRNA-seq) data for transcriptome-wide network inference.
• Comparative Analysis: Detects differentially regulated genes and gene expression signatures by comparing scGRNs across samples or conditions.
• Validation: Validated on simulated data and applied to real datasets from mouse and human sources.
• Perturbation Contexts: Has identified regulatory changes associated with acute morphine treatment, anticancer drugs, gene knockouts, double-stranded RNA stimuli, and amyloid-beta plaques across different cell types.

Scientific Applications:

• Regulatory Dynamics: Characterizes changes in gene regulatory networks underlying pathophysiological processes and responses to environmental or experimental perturbations.
• Single-Cell Mechanism Discovery: Dissects mechanisms of transcriptional regulation at single-cell resolution in mouse and human cell types.
• Perturbation Profiling: Profiles network-level responses to treatments and stimuli such as drugs, genetic knockouts, dsRNA, and amyloid-beta.
• Comparative Studies: Enables direct comparison of scGRNs from different samples or conditions to reveal shifts in gene expression programs.

Methodology:

Computational methods explicitly include principal component regression, low-rank tensor approximation, and manifold alignment.