Development of Data Processing Methods in Chemical Isotope Labeling Liquid Chromatography-Mass Spectrometry-Based Metabolomics

• Author / Creator

  Li, Yunong

• Metabolomics is an active research field on the methods development for the analysis of small metabolites in biological systems. It provides powerful approaches that allow us to examine the variations in the metabolic profiles and is capable of detecting complex biological changes using statistical pattern recognition methods. In metabolomics analysis, large amounts of data are produced routinely in order to characterize a sample consisting of hundreds to thousands of metabolites. The conclusions drawn from the metabolomics data rely on the coverage of the detection method, the accuracy of the metabolite concentrations, and the completeness of data to include all the metabolite signals. Therefore, a number of challenges are associated with the data processing methods specific to each experimental platform. The LC-MS technique has been used widely in the application of metabolomics due to its high sensitivity and high throughput. Traditional LC-MS platforms are limited by the coverage of the detection and the less reproducible quantification results. Thus, the chemical isotope labeling LC-MS method was developed in our group for an improvement of the metabolite separation and a higher detection sensitivity of a broad range of metabolites in a biological sample. In the labeling LC-MS method, each labeled metabolite will generate a peak pair signal in their mass spectra, with the light peak from the individual sample and the heavy peak from the pooled sample. Accordingly, a customized data processing method is required in dealing with the data generated by different chemical isotope labeling LC-MS experiments. My research focused on the development of data processing methods to address the challenges from the growing data processing tasks in the chemical isotope labeling LC-MS. I developed an integrated data processing workflow that checked the LC-MS raw data in terms of the mass accuracy and retention time reproducibility (Chapters 2 and 3), aligned the peak pair data from individual files, and removed the false peak pairs and redundant peak pairs to improve the confidence of each peak pair in representing a true labeled metabolite (Chapter 4). A missing ratio imputation method was developed to fill all missing ratios and generate a complete metabolite-intensity table for the statistical analyses (Chapter 4). In the peak pair ratio calculations, I developed a data processing method that accounts for natural isotope contributions in the peak intensity of the 13C2-labeled peak and improved the quantification accuracy (Chapter 5). An intensity-dependent mass tolerance method was developed to assist the mass-based database search (Chapter 6). All the processing methods were integrated in a program with a graphical user interface that has been implemented in the lab for routine data processing tasks. I used an application of a wine metabolomics study to demonstrate the data processing workflow (Chapter 7). The integrated data processing program can be applied in different chemical isotope labeling LC-MS experiments to facilitate the qualitative and quantitative analyses of different metabolomes.

• Subjects / Keywords

  • LC-MS
  • Metabolomics
  • Data processing methods
  • Chemical isotope labeling

• Graduation date

  Spring 2019

• Type of Item

  Thesis

• Degree

  Doctor of Philosophy

• DOI

  https://doi.org/10.7939/r3-3kkh-fa68

• License

  Permission is hereby granted to the University of Alberta Libraries to reproduce single copies of this thesis and to lend or sell such copies for private, scholarly or scientific research purposes only. Where the thesis is converted to, or otherwise made available in digital form, the University of Alberta will advise potential users of the thesis of these terms. The author reserves all other publication and other rights in association with the copyright in the thesis and, except as herein before provided, neither the thesis nor any substantial portion thereof may be printed or otherwise reproduced in any material form whatsoever without the author's prior written permission.