Unveiling gene perturbation effects through gene regulatory networks inference from single-cell transcriptomic data.

Area of Science:

• Genomics
• Systems Biology
• Computational Biology

Background:

• Gene regulatory networks (GRNs) govern cellular functions.
• Reconstructing GRNs aids in modeling cell behavior and predicting genetic change impacts.
• Single-cell RNA sequencing (scRNA-seq) provides rich data for GRN inference.

Purpose of the Study:

• To develop an unsupervised machine learning framework, IGNITE, for inferring GRNs from scRNA-seq data without prior knowledge.
• To enable prediction of gene perturbations using inferred GRNs.
• To advance the understanding of biological processes and key genes.

Main Methods:

• Developed IGNITE, an unsupervised machine learning framework based on the inverse problem for a kinetic Ising model.
• Inferred directed, weighted, and signed GRNs directly from unperturbed scRNA-seq data.
• Used inferred GRNs to simulate gene expression data upon single and multiple genetic perturbations.

Main Results:

• IGNITE accurately inferred GRNs from scRNA-seq data across different species and technologies.
• Simulated gene knockouts using IGNITE showed predictions consistent with experimental observations.
• IGNITE accurately predicted the effects of perturbations in murine and human pluripotent stem cells.

Conclusions:

• IGNITE robustly captures gene interaction logic from scRNA-seq data.
• The framework enables reliable in silico perturbation analyses.
• IGNITE has significant potential for understanding biological processes and identifying key genes.