An Empirical Bayes Model for Metabolite Identifications Using Two-Dimensional Gas Chromatography Mass Spectrometry

• Activity Number: 242
• Type: Contributed
• Date/Time: Monday, August 2, 2010 : 2:00 PM to 3:50 PM
• Sponsor: Section on Bayesian Statistical Science
• Author(s): Jaesik Jeong*+ and Changyu Shen and Xiang Zhang
• Companies: Indiana University Purdue University Indianapolis and Indiana University Purdue University Indianapolis and University of Louisville
• Keywords: metabolomics ; expectation-maximization ; similarity score ; GCGC/TOF-MS

Abstract

Two dimensional gas chromatography - mass spectrometry (GCxGC/TOF-MS) is an emerging technology that offers significant advantages for the analysis of metabolites present in complex samples. It offers an order-of-magnitude increase in separation capacity over one dimensional gas chromatography, leading to significant improvement in mass spectral de-convolution for metabolite identification through comparing the experimental spectrum with a library of spectra with known identities and assigning the best match. Nevertheless, the identifications of metabolites that generate the large amount of spectra are still subject to errors. Therefore, statistical/computational approaches to improve the accuracy of the identifications and validity of false positive control/estimate are in great need. We propose a hierarchical statistical model in the empirical Bayes framework to tackle this problem.