Advancing computational methods for mass spectrometry-based proteomics, metabolomics, and analysis of multi-omics datasets

Hamzeiy, Hamid (2021): Advancing computational methods for mass spectrometry-based proteomics, metabolomics, and analysis of multi-omics datasets. Dissertation, LMU München: Fakultät für Chemie und Pharmazie

Abstract

Undoubtedly, the current century is witness to an unprecedented speed in advancements within biological sciences, which are owed to the immense technological progress in the analytical tools and methods utilized, and to the dawn of the fast developing fields of omics and bioinformatics. Omics allows the collection of holistic data on several different biomolecule classes, and bioinformatics makes it possible to explore and understand the vast amounts of data produced. The most mature omics fields, in terms of both hardware and software, are genomics and transcriptomics, based on next generation sequencing (NGS) technologies. With the introduction of electrospray ionization and high-resolution mass spectrometry, liquid chromatography coupled to tandem mass spectrometry (LC-MS/MS), has made significant leaps for the fields of metabolomics and proteomics. One promising method for LC-MS/MS-based proteomics is data independent acquisition (DIA), which requires advanced data analysis algorithms. MaxDIA, within the MaxQuant software for the processing of LC-MS/MS-based proteomics data, is introduced here. It comes with an accurate false discovery rate estimation of the peptide and protein identification based on measured and predicted spectrum libraries. When compared to the state of the art, MaxDIA also delivers comprehensive proteome coverages and lower coefficients of variation in protein quantification. Bioinformatics tools for the analysis of metabolomics data generally follow the same principles and steps as proteomics software, but due the huge numbers of metabolites and immense complexity of metabolomics data, much work is still needed to bring metabolomics software to the level of maturity of their proteomics equivalents. MaxQuant is a time tested and widely accepted software for the processing of proteomics data, which was first recognized for its cutting-edge nonlinear recalibration for reaching superior precursor mass accuracy, which helps significantly improve peptide identifications. Here, following this direction, a new algorithm within MaxQuant for improving mass accuracy in metabolomics data is introduced, which utilizes a novel metabolite library-based mass recalibration algorithm. The many types of omics data available today present a great opportunity for developing approaches to combine such data in order to infer new knowledge, often termed multi-omics studies. A robust approach to this end is to utilize prior knowledge on the relationships of the various major biomolecules in question, which are often depicted in network structures where the nodes of the network depict biomolecules and the edges correspond to an interaction. To implement this approach, Metis is introduced, a new plugin for the Perseus software aimed at analyzing quantitative multi-omics data based on metabolic pathways. This thesis includes four publications, the first of which is a review article on computational metabolomics as a part of the introduction, listed below: 1. Hamzeiy, Hamid, and Jürgen Cox. 2017. “What Computational Non-Targeted Mass Spectrometry-Based Metabolomics Can Gain from Shotgun Proteomics.” Current Opinion in Biotechnology 43: 141–46. https://doi.org/10.1016/j.copbio.2016.11.014. 2. Sinitcyn, Pavel, Shivani Tiwary, Jan Rudolph, Petra Gutenbrunner, Christoph Wichmann, Şule Yllmaz, Hamid Hamzeiy, Favio Salinas, and Jürgen Cox. 2018. “MaxQuant Goes Linux.” Nature Methods 15 (6): 401. https://doi.org/10.1038/s41592-018-0018-y. 3. Pavel Sinitcyn, Hamid Hamzeiy, Favio Salinas Soto, Daniel Itzhak, Frank McCarthy, Christoph Wichmann, Martin Steger, Uli Ohmayer, Ute Distler, Stephanie Kaspar-Schoenefeld, Nikita Prianichnikov, Şule Yılmaz, Jan Daniel Rudolph, Stefan Tenzer, Yasset Perez-Riverol, Nagarjuna Nagaraj, Sean J. Humphrey and Jürgen Cox. “MaxDIA enables highly sensitive and accurate library-based and library-free data-independent acquisition proteomics.” Submitted to Nature Biotechnology, 2020 4. Hamid Hamzeiy, Daniela Ferretti, Maria S. Robles, and Jürgen Cox. “Perseus plugin ‘Metis’ for metabolic pathway-centered quantitative multi-omics data analysis supporting static and time-series experimental designs.” Submitted to Cell Systems, 2021

• Dokumententyp: Dissertationen (Dissertation, LMU München)
• Themengebiete: 500 Naturwissenschaften und Mathematik ; 500 Naturwissenschaften und Mathematik > 540 Chemie
• Fakultäten: Fakultät für Chemie und Pharmazie
• Sprache der Hochschulschrift: Englisch
• Datum der mündlichen Prüfung: 11. Februar 2021
• 1. Bericht­erstatter:in: Turck, Christoph