Consistency analysis and improvement of matabolic databases for the integration of metabolic models
Year: 2015
Type: dissertation
Abstract: Modern systems biology uses analysis and modeling of large metabolic networks. These models can be assembled by integrating data from sources SBML files and online databases. Sometimes this integration can be challenging, as information can be hidden in human-readable texts or annotational layers not directly accessible with the help of common methods. Here, it is shown how algebraic analysis can be used to unravel structural information hidden in the kinetic laws of SBML models. Additionally, this work will demonstrate the Organization Theory approach and its application for inconsistency detection on the Biomodels Database. The usefulness of combining algebraic analysis and OT is shown by comparing the gathered results with data originating from other methods, like FBA. It is shown how scientific methods can be prone to an incorrect interpretation of the data given as well as their representation format. Complementing the analysis of data given in SBML model files, we also present a tool designed to help identifying microbial communities suited to perform biodegradation tasks. The preliminaries needed to perform such a task are discussed together with problems that hinder automatic solution of metabolic research questions. Problems usually occuring in the work with databases are specified and investigated in dato; using the KEGG databse as the main source. Strategies to circumvent the found problems by rule-based network descriptions are sketched out. A detailed description of the idea of rule-based databases for metabolic and biological data will be given. Subsequently, possible applications are listed, giving examples for reasonably simple models. A new formalism will be presented which might suit the task better than more general formalisms like BGNL, which is indeed a very powerful, yet rather tedious methodology. Finally, we will give an account of the advantages and challenges of networks modeled with the rule-based description introduced in this work.
Author: Stephan Richter
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W1428888319
