Publication
Biotechnology promises the biologically and ecologically sustainable production of commodity chemicals, biofuels, pharmaceuticals and other high-value products using industrial platform microorganisms. Metabolic engineering plays a key role in this process, providing the tools for targeted modifications of microbial metabolism to create efficient microbial cell factories that convert low value substrates to value-added chemicals. Engineering microbes for the bioproduction of chemicals has been practiced through three different approaches: (i) optimization of native pathways of a host organism; (ii) incorporation of heterologous pathways in an amenable organism; and finally (iii) design and introduction of synthetic pathways in an organism. So far, the progress that has been made in the biosynthesis of chemicals was mostly achieved using the first two approaches. Nevertheless, many novel biosynthetic pathways for the production of native and non-native compounds that have potential to provide near-theoretical yields and high specific production rates of chemicals remain yet to be discovered. Therefore, the third approach is crucial for the advancement of bio-based production of value-added chemicals. We need to fully comprehend and analyze the existing knowledge of metabolism in order to generate new hypotheses and design de novo pathways. In this thesis, through development and application of efficient computational methods, we took the research path to expand our understanding of cell metabolism with the aim to discover novel knowledge about metabolic networks. We analyze different aspects of metabolism through five distinct studies. In the first study, we begin with a holistic view of the enzymatic reactions across all the species, and we propose a computational approach for identifying all the theoretically possible enzymatic reactions based on the known biochemistry. We organize our results in a web-based database called â Atlas of biochemistryâ . In the second study, we focus on one of the most structurally diverse and ubiquitous constituents of metabolism, the lipid metabolism. Here we propose a computational framework for integrating lipid species with unknown metabolic/catabolic pathways into metabolic networks. In our next study, we investigate the full metabolic capacity of E. coli. We explore computationally all enzymatic potentials of this organism, and we introduce the â Super E. coliâ , a new and advanced chassis for metabolic engineering studies. Our next contribution concentrates on the development of a new method for the atom-level description of metabolic networks. We demonstrate the significance of our approach through the reconstruction of atom-level map of the E. coli central metabolism. In the last study, we turn our focus on studying the thermodynamics of metabolism and we present our original approach for estimating the thermodynamic properties of an important class of metabolites. So far, the available thermodynamic p