Deciphering living networks : Perturbation strategies for functional genomics
Fuente van Bentem, A. de la
MetadataShow full item record
Thesis: Deciphering living networks: Perturbation strategies for functional genomics Alberto de la Fuente email@example.com Molecular Cell Physiology Free University Amsterdam Advisors: Prof.Dr. H.V. Westerhoff Prof.Dr. J.L. Snoep Supervisor: Dr. P.J. Mendes Using modern experimental techniques it is possible to measure the concentrations of a great many, and ultimately all, cellular constituents such as mRNAs, proteins and metabolites. Given these experimental technologies, astronomical amounts of new data will appear. To enable us to see the forest for the trees, we need to find ways in which best to analyze the data so as to obtain better understanding of biochemical systems and predictive power. When those new ways of analyzing the data are found, this may even lead to a preference for a certain type of data or certain experimental methodologies. This may then help direct experimentation towards the highest possible impact for understanding of biochemical systems. Ideally, the three levels of biochemical organization, i.e. mRNAs, proteins and metabolites, are studied all together in an integrated fashion. However, due to the number of components and complexity of such integrated systems it is reasonable to try to decompose the system and to study the subsystems or to use simplified descriptions of the whole system. It will be important to decompose the system into subsystems that behave in isolation in much the same way as they do when they are embedded in the whole system. This is exactly what I deal with in my dissertation; on the one hand I show how and when it is possible to study the systems properties of metabolism in vivo, ignoring the effects of gene and protein expression, and on the other hand I develop a quantitative concept in terms of Metabolic Control Analysis to describe the properties of the whole system in a simplified form, i.e. as a gene network a description of only the dynamics of gene expression without explicit accounting for metabolites and proteins. This concept enables the inference of the topology of such gene networks from experimental data. The analysis guides the experimenter towards the specific experiments that need to be done in order to be able to infer the interactions between genes on a genome scale. After introducing the relevant preliminaries in Chapter 1, in Chapter 2 I introduce the concept of hierarchical biochemical systems and show how to express their properties in terms of properties of the individual flux-disconnected modules of which it is composed. In particular, I focus on the study of metabolic systems. I propose several methods with the goal of distinguishing regulation that takes place at the metabolic level only from regulation that involves transcription or translation, thus quantifying the relative importance of each of these processes to the global systems behavior. I verify the experimental applicability of these methods by analyzing data obtained by simulation of a biochemical system. In Chapter 3 I introduce the concept of the gene network. Gene networks are network models in which the nodes represent gene activities (mRNA levels) and the edges correspond to regulatory interactions between them. Such models are highly phenomenological because they do not represent explicitly the proteins and metabolites that mediate those interactions. I show the use of Regulatory Strengths to quantify gene-gene interactions and show how to express these coefficients in terms of the biochemical system underlying these interactions. This approach establishes a clear and formal link between the phenomenological gene network modeling and more detailed approaches considering the hierarchical structuring of biochemical networks as introduced in Chapter 2.