高寒环境失血性休克大鼠代谢特征及诊断相关生物标志物:基于血浆代谢组学和3种机器学习算法的探索性研究OA

Objective To investigate the metabolic characteristics of rats with hemorrhagic shock in a high-altitude cold environment using plasma metabolomics,and to screen for potential biomarkers applicable to the diagnosis of high-altitude cold combined with hemorrhagic shock.Methods A total of 140 Sprague-Dawley rats were randomly divided into control group and high-altitude cold environment combined with uncontrolled lethal hemorrhagic shock(H-UHS)group(n=70).After both groups were exposed to-15℃for 12 h,an uncontrolled hemorrhagic shock model was established in H-UHS group.Vital signs,blood gas parameters,blood biochemistry,and coagulation indices were measured.Plasma metabolomic profiling was performed using ultra-high-performance liquid chromatography-tandem mass spectrometry(UHPLC-MS/MS).Differentially abundant metabolites were screened by three-dimensional unsupervised principal component analysis(3D-PCA),orthogonal partial least squares discriminant analysis(OPLS-DA),and weighted gene co-expression network analysis(WGCNA).Three machine learning algorithms[least absolute shrinkage and selection operator(LASSO),support vector machine with recursive feature elimination(SVM-REF),and random forest(RF)]were further applied in a cross-validation framework to identify characteristic metabolites,and their diagnostic performance was evaluated using receiver operating characteristic(ROC)curves.Results Compared with control group,H-UHS group exhibited significantly lower mean arterial pressure(MAP),heart rate(HR),and respiratory rate(RR)along with elevated blood lactate,decreased bicarbonate,significantly increased hepatic enzymes[alanine transaminase(ALT),aspartase aminotransferase(ALT)],cardiac troponin T,and D-lactate,and prolonged coagulation time,all difference are statistically significant(P<0.05).Metabolomic analysis identified 1470 differential metabolites,which were predominantly enriched in pathways such as fructose and mannose metabolism,steroid hormone biosynthesis,and secondary bile acid biosynthesis.Four characteristic metabolites:prephenic acid,selenocysteine,2'-deoxyuridine-5'-diphosphate,and phosphoethanolamine were consistently selected by all three machine learning algorithms.Their areas under the ROC curves(AUCs)for discriminating high-altitude cold hemorrhagic shock were 0.945,0.979,0.953,and 0.991,respectively.Conclusion Prephenic acid,selenocysteine,2'-deoxyuridine-5'-diphosphate,and phosphoethanolamine may serve as potential diagnostic biomarkers for high-altitude cold combined with hemorrhagic shock.