A gene regulatory network-aware graph learning method for cell identity annotation in single-cell RNA-seq data.

Area of Science:

• Genomics
• Computational Biology
• Bioinformatics

Background:

• Single-cell transcriptome data analysis is vital for cell atlases and disease research.
• Current methods struggle with data from diverse sources (batches, technologies, tissues, species).
• Gene regulatory relationships are robust across these variations, offering a stable basis for analysis.

Purpose of the Study:

• To develop an automated tool, scHGR, for robust cell identity annotation.
• To leverage gene regulatory relationships for cell communication graph construction.
• To improve noise reduction and identify distant cellular connections in single-cell data.

Main Methods:

• scHGR utilizes gene regulatory networks to build gene-mediated cell communication graphs.
• The tool is designed for automated annotation of single-cell transcriptome data.
• Benchmarking was performed across 22 diverse experimental scenarios.

Main Results:

• scHGR demonstrated precise and consistent cell identity annotation, outperforming existing methods.
• Novel subtypes within CD4+ T cells and cytotoxic T cells were identified.
• Analysis of a 56-cell type atlas for COVID-19 patients revealed key factors like IL1 and calcium ions.

Conclusions:

• scHGR provides a robust and accurate method for single-cell data annotation by leveraging gene regulatory information.
• The tool effectively handles data heterogeneity and uncovers biologically significant cell subtypes and communication pathways.
• Findings offer potential insights for targeted therapeutic interventions, particularly in the context of infectious diseases like COVID-19.