Exploring Immune-Related Gene Expression Up To The First 24-Hour For Predicting Sepsis Outcomes Based On Comprehensive Bioinformatics Analysis And Machine Learning

Abstract

Background: Host immune dysregulation participates in the prognosis of sepsis with high morbidity and mortality. Our study aimed to identify the roles of immuneassociated genes during sepsis progression and to predict sepsis survival up to 24 h at diagnosis, which may help plan future individualized treatments. Methods: GSE54514, GSE57065, and GSE95233 datasets were downloaded from the Gene Expression Omnibus (GEO) database for early identification of differentially expressed IRGs between sepsis patients and healthy controls. Candidate IRGs significantly associated with sepsis survival were obtained by univariate logistic regression analysis. Gene signatures of these IRGs were further selected by Least Absolute Shrinkage and Selection Operator (LASSO) regression and Random Forest Algorithm (RFA). The correlation between signature genes and prognosis was analyzed.Furthermore, signature IRGs were further validated by quantitative PCR (qPCR) on the whole blood of septic patients and an external COVID-19 dataset and candidate drug were predicted. Results: HLA.DPA1, IL18RAP, MMP9, RNASE3, S100P, and PTX3 were found significantly differentially expressed starting very early after sepsis infection and persisting for up to 5 days, and their formed IRG score had a satisfactory predictive value on sepsis outcome. Furthermore, our validation showed that these six IRGs were also significantly dysregulated in both an external COVID-19 dataset and sepsis patients. Finally, 10 potential compounds were predicted to have targeted these genes. Conclusion: Our study developed a prognostic modeling tool for sepsis survival based on IRG expression profiles, and has the capacity for early prediction of sepsis outcomes via monitoring the immunogenomic landscape, and possibly the individualized therapies for sepsis survival.