Neutrophil-based single-cell sequencing combined with transcriptome sequencing to explore a prognostic model of sepsis

Abstract Sepsis is a life-threatening condition influenced by various factors. Although gene expression profiling has offered new insights, accurately assessing patient risk and prognosis remains challenging. We utilized single-cell and gene expression data of sepsis patients from public databases. The Seurat package was applied for preprocessing and clustering single-cell data, focusing on neutrophils. Lasso regression identified key genes, and a prognostic model was built. Model performance was evaluated using Receiver Operating Characteristic (ROC) curves, and further analyses, including immune cell infiltration, Gene Set Enrichment Analysis (GSEA), and clinical correlation, were conducted. Several neutrophil subtypes were identified with distinct gene expression profiles. A prognostic model based on these profiles demonstrated strong predictive accuracy. Risk scores were significantly correlated with clinical features, immune responses, and key signalling pathways. This study provides a comprehensive analysis of sepsis at the molecular level. The prognostic model shows promise in predicting patient outcomes, offering potential new strategies for diagnosis and treatment.