Open Collections

Guo, Jian

University of British Columbia

Untargeted metabolomics studies the complete set of small metabolic molecules in a given biological system. Liquid chromatography coupled with high-resolution mass spectrometry (LC-HRMS) is currently the most prominent analytical platform for untargeted metabolomics owing to its high sensitivity, specificity, and metabolic coverage. However, the current LC-MS-based untargeted metabolomics workflow has limited performance in detecting and quantifying trace-level metabolites of bad chromatographic peak shapes. It is also hard to differentiate signals of real metabolites from noise and background. During my Ph.D., I have developed a suite of analytical and bioinformatic tools to address the critical challenges in metabolomics data acquisition and data processing. In this thesis, Chapters 2 to 4 focus on the development of data acquisition methods. Specifically, Chapters 2 to 3 describe the detailed comparison of the existing data acquisition modes in two different aspects. Chapter 4 describes the development of a novel data acquisition strategy, DaDIA, to increase metabolomic coverage. On the other hand, Chapters 5 to 9 describe the development of data processing methods. In Chapter 5, a data processing parameter optimization tool, Paramounter is introduced to rapidly and accurately determine the best peak-picking parameters for five commonly used data processing programs. In Chapter 6, the five most commonly used metabolomics data processing programs were compared to mechanistically explain the difference regarding the performance in metabolic feature extraction. In Chapter 7, a novel data processing program, JPA, was developed to efficiently extract the metabolic features by combining multiple peak-picking algorithms. In Chapter 8, a deep learning-based software, EVA, was created to automatically remove the false positive features generated from the background noise. In Chapter 9, I developed a bioinformatic workflow, ISFrag, to automatically recognize and remove the false positive metabolic features originating from in-source fragmentation.

Text

2023-04-06

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/84146

Chemistry

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/