Identification of FTO as a key m6A demethylase linking immune dysregulation to sepsis pathogenesis.

Background: Sepsis is a life-threatening disorder characterized by multiple organ dysfunction caused by dysregulated host responses to infection. The present study aimed to identify potential diagnostic biomarkers for sepsis and elucidate their molecular mechanisms through comprehensive bioinformatics and experimental analyses. Methods: Five publicly available transcriptomic datasets (GSE13904, GSE26440, GSE28750, GSE95233, and GSE57065) containing sepsis and healthy control samples were utilized in the study. After quality control and normalization, the samples were divided into training and validation cohorts. Fourteen machine learning algorithms were applied to the training cohort to identify robust diagnostic biomarkers, and their predictive performance was subsequently verified in the validation cohorts. Single-cell RNA sequencing (scRNA-seq) data were further analyzed to determine the cellular distribution of the identified regulators among immune cell subsets. Results: In total, the least absolute shrinkage and selection operator (LASSO) model exhibited the best performance in the validation set, demonstrating high reliability. Through consensus feature selection across multiple models, the m Conclusion: FTO, identified through consensus machine learning approaches, could serve as a potential diagnostic biomarker and m