Multi-Omics and Machine Learning-Based Profiling of Severity Signatures in Mycoplasma Pneumoniae Infection in Children

Mycoplasma pneumoniae pneumonia (MPP) is a common respiratory infection in children, yet the mechanisms driving its progression to severe disease remain poorly understood. This study employs a comprehensive proteomic and metabolomic approach to elucidate severity-related pathways and identify potential biomarkers for improved diagnosis and targeted therapy. By analyzing blood proteomes from 57 pediatric patients with varying MPP severities alongside 10 healthy controls, and integrating multi-omics data from bronchoalveolar lavage fluid (BALF), we uncovered key severity-associated proteins and metabolites linked to inflammatory and metabolic dysregulation. Notably, alterations in L-arginine metabolism, influenced by APAF1 and SERPINB5, were found to modulate the efferocytosis pathway, while ERCC1 and MLH1, crucial components of the Fanconi anemia pathway, were associated with changes in plasmid acid metabolites affecting fatty acid elongation. Machine learning analysis further identified three critical biomarkers—TNFRSF10B, SAT1, and 4-nitrophenol— that accurately distinguished between mild and severe MPP cases with high sensitivity (0.8) and specificity (1). These findings provide novel insights into the molecular mechanisms underlying severe MPP, highlighting efferocytosis and fatty acid metabolism as key pathways. The identification of severity-specific biomarkers offers a foundation for enhanced diagnostic precision, improved disease stratification, and the development of targeted therapeutic strategies to optimize the management of severe MPP in pediatric patients.