scPriorGraph: constructing biosemantic cell–cell graphs with prior gene set selection for cell type identification from scRNA-seq data

• Published in: Genome Biology Vol. 25; no. 1; p. 207
• Main Authors: Cao, Xiyue, Huang, Yu-An, You, Zhu-Hong, Shang, Xuequn, Hu, Lun, Hu, Peng-Wei, Huang, Zhi-An
• Format: Journal Article
• Language: English
• Published: England BioMed Central 05.08.2024; BMC
• Subjects: Algorithms; Cell-type identification; Classification; Communication; Computational Biology - methods; computational methodology; Data analysis; Datasets; Dual-channel graph neural network; Gene expression; genes; Graph augmentation; Graphs; Humans; Ligand-receptor network; Ligands; Machine learning; Medical research; Metabolism; Method; Neural networks; Neural Networks, Computer; Noise reduction; Pathway; RNA-Seq - methods; Semantics; Signal transduction; Single-Cell Analysis - methods; Single-Cell Gene Expression Analysis; Single-cell RNA sequencing; Software; Sparsity; Single-cell RNA sequencing; Cell-type identification; Dual-channel graph neural network; Ligand-receptor network; Pathway; Graph augmentation
• ISSN: 1474-760X; 1474-7596; 1474-760X
• DOI: 10.1186/s13059-024-03357-w

Cell type identification is an indispensable analytical step in single-cell data analyses. To address the high noise stemming from gene expression data, existing computational methods often overlook the biologically meaningful relationships between genes, opting to reduce all genes to a unified data space. We assume that such relationships can aid in characterizing cell type features and improving cell type recognition accuracy. To this end, we introduce scPriorGraph, a dual-channel graph neural network that integrates multi-level gene biosemantics. Experimental results demonstrate that scPriorGraph effectively aggregates feature values of similar cells using high-quality graphs, achieving state-of-the-art performance in cell type identification.