Lingjue Wang


Date of Award

Spring 5-15-2023

Author's School

Graduate School of Arts and Sciences

Author's Department


Degree Name

Doctor of Philosophy (PhD)

Degree Type



Untargeted Metabolomics aims at comprehensive measurement of small molecules termed metabolites that involve in biochemical processes such as energy production, biomass production, and antioxidation. The use of mass spectrometry (MS) coupling to separation technology such as liquid chromatography (LC) and ion mobility spectrometry (IMS) serves as the most prevailing approach to perform untargeted metabolomics. While it is relatively routine to generate tens of thousands of signals commonly referred as features from LC-MS data acquisition, inference of signals derived from the biological samples remains the major challenges in untargeted metabolomics workflow. To overcome this challenge, major efforts were made in my thesis work with the develop of three informatic tools for untargeted metabolomics as well as a LC-MS assay in targeted metabolomics. In the first work, I developed a stable isotope-based protocol called credentialing to compare analytical methods for untargeted metabolomics. In this workflow, I detailed how to prepare credentialed E. coli samples. I upgraded the original credential R package to version 3.1, which incorporates a new algorithm to detect isotopologue counterparts from a format-neutral feature table, expanding the compatibility of this workflow to a broader range of data processing pipelines. In the second work, I developed TOXcms software that incorporates dose-response analysis in untargeted metabolomic workflow. I demonstrated the utility of dose-response metabolomics analysis to understand biochemical mechanisms and identify off-target effects of small molecule drugs. In the third work, I developed XCiMS software to process four-dimensional (4D) LC-IMS-MS data. XCiMS was built within the XCMS ecosystem and supports data acquired from three major IMS-MS instrument platforms including traveling wave (TWIMS), drift tube (DTIMS), and trapped ion mobility spectrometry (TIMS). XCiMS detects 4D unique features and sub-features from the raw data with three molecular descriptors: mass-to-charge ratio (m/z), retention time, and ion mobility-derived collision cross-section (CCS). Using XCiMS, I processed 11 datasets acquired from three IMS-MS platforms and analyzed compositions of 4D sub-features. Other than IMS-resolved structural isomers, two new events of ion degeneracy including charge isomers and cluster ion dissociation were identified. I also demonstrated that CCS, as a complementary molecular descriptor, facilitated metabolite annotation by reducing false positive hits in metabolite database. Together, I expanded the toolkit for 4D untargeted metabolomics and uncovered complex ion compositions by IMS. In the fourth work, I developed a fast and standard-free LC-MS assay to quantify NADPH/NADP+ and NADH/NAD+ ratios in cellular and tissue samples. MS response calibration factors for each pairs redox cofactors were established in three mass-spectrometers as empirical constants to calculate the actual ratios from MS responses directly. The assay was cross-validated with conventional LC-MS assay with standard calibration curves and colorimetric assay. Using this LC-MS assay, I successfully quantified ratios of these redox cofactors in a panel of cancer cell lines. This thesis constitutes a body of work with novel advance in methodologies for untargeted and targeted metabolomics that enables detection of biologically relevant information from complex metabolomics data. Collectively, this work facilitated novel discovery of the ion composition in untargeted metabolomics data as well as understanding of biochemical mechanisms.


English (en)

Chair and Committee

Gary J. Patti Michael Gross

Committee Members

Michael R. Brent, Joseph A. Fournier, Timothy A. Wencewicz,

Available for download on Friday, May 10, 2030