基于OPLS—DA的高脂血症血浆1H-NMR研究

1H-NMR-based metabonomics research on plasma from patients with hyperlipidemia with OPLS-DA

目的:分析高脂血症患者与正常人血浆内源性代谢指纹图谱变化,探讨代谢组学在高脂血症研究中的应用。方法:应用核磁共振氢谱(1H-NMR)仪检测正常对照(C)组和高脂血症(HLP)组各10例患者血浆标本得到代谢指纹图谱,采用正交偏最小二乘判别分析(OPLS-DA)进行模式识别,并通过模型分析和非参数检验鉴别血浆代谢产物的变化及潜在生物标志物。结果:C组和HLP组代谢产物有明显不同,两组代谢产物各自聚集,分布区域明显分开。与C组比较,HLP组LDL/VLDL、乳酸、糖蛋白水平明显增加,而丙氨酸、缬氨酸、胆碱、肌酸、酪氨酸、谷氨酰胺及血糖水平下降。模型主成分回归系数R2=0.885,Q2=0.332。筛选到30个变量13个代谢产物可能是高脂血症潜在生物标志物。结论:高脂血症患者血浆代谢组学发生明显变化,可通过1H-NMR发现高脂血症的潜在生物标志物。

Objective To explore the changes of metabolic finger spectra from plasma of patients with hyperlipidemia（I-ILP） and investigate the application of metabonomic. Methods The metabolic finger spectra was investigated using 1H-NMR spectroscopy of plsama combined with pattern recognition techniques orthogonality partial least squares-discriminant analysis （OPLS-DA） and nonparametric test to detect differences between groups and the biomarkers. Results The OPLS-DA plots of the plasma samples presented marked clustering between control group and HLP group.Compared to control group the level of low density lipoprotein （LDL）/very low density lipoprotein （VLDL）,lactate andglycoprotein were higer while alanine （Ala）,valine （Val）, choline,creatine,tyrosine （Tyr）, glutaminate and glucose were lower. Conclusion 1H-NMR-based OPLS-DA model can diagnose hyperlipidemia. Obvious changes in metabonomics of plasma from hyperlipidemia were observed and potential biomarkers can be detected by 1H-NMR.