Abstract
Metabolomics is a useful approach to explore systemic metabolic variation and to elucidate disease mechanisms. In this study, human plasma metabolic profiles of coronary heart disease (CHD) patients and healthy controls were obtained by gas chromatography-mass spectrometry (GC-MS). A relatively new pattern recognition method, the Monte Carlo tree (MCTree) approach, was used to explore metabolic differences between CHD patients and healthy controls. In this way, CHD patients with different severity of coronary atherosclerosis were classified by the corresponding metabolic profiles. Furthermore, important metabolites contributing to the classification were screened and identified by their mass spectra. Several potential biomarkers were discussed in some detail. The results demonstrated that the proposed method might be a useful tool for discovering metabolic abnormalities and potential biomarkers for diseases.
Acknowledgments
This work was supported financially by National Nature Foundation Committee of P. R. China (No. 21105129 and No. 21075138), Special Foundation of China Postdoctoral Science (No. 200902481), and the Fundamental Research Funds for the Central Universities (2010QZZD010 and 2011QNZT053).
Notes
Note: Data are presented as means ± SD.
Note: data are presented as means ± SD by 100 times the ratio of its peak area to that of the internal standard on the same chromatogram.