Identification and Validation of Mitochondria-Related Genes for Diagnosis of Early-Stage Sepsis.

OBJECTIVE

Sepsis is a life-threatening condition with unclear pathogenesis and limited effective treatments. Mitochondrial dysfunction is considered a key factor in sepsis-induced multiple organ failure. This study aimed to identify essential mitochondria-related genes associated with sepsis to improve diagnosis and treatment strategies.

METHODS

High-throughput gene expression data (GSE185263) were analyzed to identify differentially expressed genes (DEGs) in 348 septic patients and 44 healthy controls. Mitochondria-related DEGs were screened using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases. Two machine-learning algorithms, LASSO and SVM-RFE, were applied to identify mitochondria-associated hub genes.

RESULTS

We identified 548 DEGs and screened 18 mitochondria-related DEGs. LASSO and SVM-RFE analyses identified 11 genes associated with sepsis diagnosis, showing strong diagnostic abilities through ROC assays. The expression of these 11 genes was examined by quantitative real-time polymerase chain reaction in septic patients and healthy participants, and differential expression of arginase 2 (ARG2), B-cell lymphoma 2-related protein A1 (BCL2A1), interferon alpha inducible protein 27 (IFI27), NADH: ubiquinone oxidoreductase subunit B3 (NDUFB3), stomatin (STOM), and translocator protein (TSPO) were observed. Some gene expression differences remained significant after adjusting for neutrophil and platelet counts.

CONCLUSIONS

These findings suggest that mitochondrial dysfunction plays a critical role in sepsis progression, and the identified genes may serve as biomarkers for early diagnosis and targeted treatment, potentially improving patient outcomes.