Identification of Biomarkers and Construction of Predictive Models for Sepsis and Septic Shock

Sepsis is a serious condition resulting from an uncontrolled immune response to infection, often leading to organ dysfunction and septic shock. Current biomarkers have limitations in reflecting disease severity. Recent advances in gene expression analysis suggest that identifying novel biomarkers could improve early diagnosis and intervention, potentially enhancing patient outcomes in sepsis and septic shock.

A whole blood RNA-Seq dataset was obtained from the public database, consisting of samples from sepsis, septic shock, and healthy control groups. Hub genes were identified using differential expression analysis and weighted gene co-expression network analysis. Functional analysis was performed using Gene Ontology and Gene Set Enrichment Analysis. Expression levels of each hub gene across different groups were compared. Biomarkers were identified and predictive models for sepsis and septic shock were constructed using stepwise regression and logistic regression. The models were validated using external datasets.

Nine hub genes were identified, with expression levels showing an upward trend in sepsis and septic shock samples. These hub genes were enriched in pathways related to the innate immune system and neutrophils. Predictive models for sepsis (with participating biomarkers ELANE, OLFM4, and MMP8) and septic shock (with participating biomarker COL17A1) demonstrated good diagnostic efficacy during validation.

This study identified biomarkers and developed predictive models for early identification of sepsis and septic shock, which could improve patient prognosis. Further investigations are needed to understand the underlying mechanisms of these biomarkers in sepsis.