Statistical Analysis

We have developed advanced statistical methods for analyzing metabolomics data. We developed ratio analysis approaches (Ratio Analysis of Mass Spectrometry) to improve unknown identification in metabolomics by reducing the interference from other, unrelated metabolites (chemical noise). In addition, we combined NMR and MS in various ways, and we found that NMR and MS can generate different metabolic profiles from a sample; thus, their combination can be very valuable in metabolomics, not only because one can thereby obtain a more comprehensive profile of detectable metabolites, but also because the reliability and predictive accuracy of statistical models can be potentially improved.

1. H. Gu, Z. Pan, B. Xi, V. Asiago, B. Musselman, D. Raftery*, “Principal Component Directed Partial Least Squares Analysis for Combining NMR and MS Data in Metabolomics: Application to the Detection of Breast Cancer”, Anal. Chim. Acta, 686, 57-63, 2011. PMCID: PMC3040422

2. L. Deng⁺, H. Gu⁺*, J. Zhu, G. A. Nagana Gowda, D. Djukovic, E. G. Chiorean, and D. Raftery*, “Combining NMR and LC/MS Using Backward Variable Elimination: Metabolomics Analysis of Colorectal Cancer, Polyps, and Healthy Controls”, Anal. Chem., 88, 7975-7983, 2016. PMID: 27437783