Predicting gene regulatory networks from cell atlases

Area of Science:

• Computational biology
• Genomics
• Systems biology

Background:

• Single-cell RNA sequencing (scRNA-seq) has enabled comprehensive cell type identification across mouse tissues.
• Understanding gene regulatory networks is crucial for deciphering cell identity and function.
• Technical variations in scRNA-seq data present challenges for integrated analysis.

Purpose of the Study:

• To computationally reconstruct gene regulatory networks from multiple mouse scRNA-seq atlases.
• To identify global and cell type-specific gene regulatory modules (regulons).
• To investigate the functional role of regulons in cell identity and differentiation.

Main Methods:

• Computational reconstruction of gene regulatory networks from existing mouse scRNA-seq atlases.
• Identification and distinction of global and cell type-specific regulons.
• Integration of networks to uncover coordinated regulon modules.
• Validation of modules using experimental data and analysis of myeloid differentiation.

Main Results:

• Gene regulatory networks accurately capture functional regulators critical for cell identity.
• Identified distinct global regulons active across multiple cell types and specialized cell type-specific regulons.
• Regulon activities effectively distinguish individual cell types despite inter-atlas technical differences.
• Uncovered coordinated regulon modules essential for cell types and validated their function.
• Elucidated the role of the Irf8 regulon in myeloid differentiation and monocyte lineage.

Conclusions:

• Computational reconstruction of gene regulatory networks from scRNA-seq data is a powerful approach to identify cell identity regulators.
• Integrated analysis of multiple atlases reveals conserved and specific regulatory principles.
• The identified regulons and modules provide insights into cellular function and differentiation, with potential applications in understanding disease.