论文部分内容阅读
目的:研究基于惩罚岭型线性判别分析(PRLDA)对乙酰氨基酚(APAP)为肝毒性模型药物,与传统的主成分分析(PCA)比较建立的代谢组学模型的方法学。方法:应用APAP对大鼠肝毒性造模,利用UPLC-MS检测尿液样本,以总离子流色谱为毒性表达变量,运用PRLDA和PCA分别建立APAP肝毒性代谢组学模型。结果:在PRLDA代谢组学模型中,APAP组明显偏离正常组,可以直观看出正常组与模型组的类间差异,具有较强的分类能力;而PCA的分类能力较差。结论:PRLDA代谢组学模型的建立能准确、灵敏地表达药物毒性,并且该方法优于PCA模型。
OBJECTIVE: To study the methodology of metabonomics models based on the comparison of penalized Ridge Linear Discriminant Analysis (PRLDA) with acetaminophen (APAP) as a hepatotoxic model drug compared with traditional principal component analysis (PCA). Methods: The hepatotoxicity of rats was induced by APAP. The urinary samples were detected by UPLC-MS. The total ion chromatogram was taken as the toxic expression variable. The APAP hepatotoxicity metabolomics model was established by PRLDA and PCA respectively. Results: In the PRLDA metabolomics model, the APAP group deviated significantly from the normal group, the difference between the normal group and the model group can be directly seen, with strong classification ability; while the PCA classification ability is poor. Conclusion: The establishment of PRLDA metabolomics model can express drug toxicity accurately and sensitively, and the method is superior to PCA model.