Abstract
Sepsis, a systemic inflammatory response to infection, remains a significant health challenge with high morbidity and mortality rates. The molecular mechanisms underlying sepsis, particularly the role of programmed cell death (PCD), are not fully understood. This study is aimed at elucidating the transcriptomic changes associated with sepsis, emphasizing PCD, and identifying potential diagnostic biomarkers. Transcriptome data from sepsis and control samples were extracted from the GEO website. Differential expression analysis identified genes perturbed in sepsis. WGCNA revealed 14 highly connected modules, with the turquoise module showing the strongest association with sepsis. A set of 262 hub genes was identified, which were mainly associated with apoptotic signaling pathways. Seven prognostic-related overlapping feature genes (PRGs) were identified. More importantly, the diagnostic model, constructed using eight machine learning algorithms, exhibited high efficacy in distinguishing sepsis patients from controls. The validation of feature genes at the scRNA-seq level adds a layer of robustness to our conclusions. The strong association of genes like S100A9 and KLHL3 with neutrophils, pivotal players in sepsis, suggests potential avenues for therapeutic targeting. Our comprehensive analysis has unveiled the significant role of PCD in sepsis. The insights gained from this study provide a foundation for future therapeutic interventions.