Abstract
At present, there are currently no molecular biomarkers for the early diagnosis of sepsis cardiomyopathy (SCM) in clinical practice. This study focuses on an in-depth examination of the DNA hydroxymethylation profiles within plasma extracellular vesicles and explores potential molecular biomarkers during the process of SCM. The 5hmC-Seal sequencing technology was utilized to examine the hydroxymethylation modifications of extracellular vesicles DNAs in 13 patients with septic cardiomyopathy, 18 patients with sepsis without cardiomyopathy, and 8 patients without sepsis. Additionally, a diagnostic model was constructed using machine learning methods based on the differential hydroxymethylation modifications to screen for candidate biomarkers. The accuracy of the diagnostic model was 0.962, with a sensitivity and specificity of 92.3% and 88.89%, respectively. Furthermore, the diagnostic accuracy was validated using the GEO dataset, with an accuracy rate reaching 1 (GSE79962 and GSE66890), and the differential diagnostic accuracy rates also reached 0.959 and 0.944 (GSE79962). Together, the results suggest that extracellular vesicles DNAs hydroxymethylation markers can be used for diagnosis of septic cardiomyopathy.