Theses
On this site you can find successful bachelor and master theses that were made under the supervision of our institute. The students made their theses/personal information available for the public on a voluntary basis. However, the author/copyright holder can revoke the use of their thesis/personal data at any time, by informing the Institiute of Quantitative and Theoretical Biology in written form. If you are interested in reading one of these theses below, please contact Prof. Dr. Oliver Ebenhöh.
Bachelor's theses
Caroline Lau: Comparing interaction models in arctic marine microbial ecosystems
The importance of microorganisms in modern biology and life sciences is undisputed, and still there are vast areas which have received poor, if any scientific attention. This relates to a large degree for marine, and even more so for questions regarding the arctic. This thesis aims to address the present gaps and offer new insights into the arctic microbial community and respective, their interactions amongst one another. For this purpose, I have analysed 6 interactions and light-cycle graphs each, which I created using Python code, the Protist Interaction Database and a dataset obtained from the FRAM project of AWI Bremerhaven. I evaluated the interaction graphs accordingly to their assigned interaction type (predatory, parasitic and symbiotic), and the light-cycle graphs with respect to the diet of the organisms (phototroph, mixotroph and heterotroph). 3 of the interaction graphs and 4 of the light-cycle graphs behaved in an expected manner, with the provided information of the interactions and diets in mind. For the remaining graphs, I attempted to find explanations, and in case of detected flaws, solutions to provide future research about this topic with an optimised workflow.
Tobias Blacha: "Funktionale Ähnlichkeit angrenzender Gene von Prophagen"
Die Arbeit mit Phagen ist heute aktueller denn je, da sie alternative Behandlungsmöglichkeiten bietet, bakterielle Infektionen von bereits antibiotikaresistenten Erregern zu behandeln. In dieser Arbeit wird ein Verfahren vorgestellt, um signifikante Proteindomänen im Bakteriengenom anhand der direkten Nachbarschaft bereits im Genom integrierter Prophagen zu ermitteln. Signifikante Domänen ähnlicher Funktionalität konnten in der Nachbarschaft verschiedener Prophagen in unterschiedlichen Bakterien nachgewiesen werden. Eine Klassikfikation der Domänen ist häufig nicht gegeben, da die Funktionen im Vergleich zu Lebewesen wie dem Menschen, aufgrund der ihrer Vielfalt nur begrenzt bekannt sind. Die Vorgestellten Ergebnisse basieren auf der Grundlage von statistischen Wahrscheinlichkeitswerten, die mit dem Fisher-Test erzeugt wurden. Mit den Ergenissen werden Aussagen über Zusammenhänge von funktionaler Ähnlichkeit einzelner Domänen getroffen.
Barbara Ezeani: "South Africa’s biomes: Analysis of photosynthetic parameters of plants in diverse environments and climates"
South Africa’s Capensis consists of eight biomes Fynbos, Succulent Karoo, Namo-Karoo, Desert, Grassland, Savanna, Albany Thicket, Indian Ocean Coastal Belt, and Forest. Biomes are habitats where flora and fauna occur through a shared climate. Due to the different climate regions within Capensis, the thesis focuses on photochemical processes with emphasises on non- photochemical processes. Therefore, basic knowledge of photosynthesis is required. The principle of photosynthesis is the conversion of light energy and carbon dioxide to glucose and oxygen. In the light reaction, the two light-harvesting complexes (LHCs) PhotosystemII(PSII) and PhotosystemI(PSI) absorb the sunlight and turn it into energy, but when these LHCs get over-excited with energy the plants can decay. Consequently, there are different pathways to protect the plant, one being non-photochemical quenching (NPQ) of PSII. The NPQ is the heat dissipation of over-excited LHCs.
The aim is to determine whether the NPQ of the plants in different biomes show deviations. The PhotosynQ device MultispeQ measures the fluorescence response at the actinic lights from 10, 50, 150 and 500 µmol ∙ s-1 ∙ m-2 from which the NPQ can be analysed. The leaves are clamped by the device that then uses the pulse amplitude modulation (PAM) method with a self-programmed schedule.
Through a summarization of the NPQ during their different actinic lights, the NPQ response beginning from 150 µmol ∙ s-1 ∙ m-2 displays that out of the five selected plants two react as expected to their biome. Those that are found in biomes with excessive light exposure. However, some biomes have high light exposure but, in the topography, the plants are protected by shaded territories.
Elena Kullmann: "Modelling of the photosynthetic electron flow in Cyanobacteria"
Even though scientific research becomes more interested in photosynthesis in cyanobacteria regarding issues like the climate change [13] and mathematical modelling becomes a more popular tool when trying to understand biological processes [3], it is difficult to find a reproducible model of the photosynthetic electron transport chain in cyanobacteria. In this thesis, a simplified model of the photosynthetic electron flow in cyanobacteria based on the process in Synechocystis sp. PCC 6803 is presented. Some first analyses with the model show its functionality even though further improvements can be made.
Alexander Zubov: "Co-Cultures as Model Systems: Phototroph-Heterotroph Interactions to Simulate Ecological Nutrient Cycles"
Ecosystems represent highly complex, large biological networks with global importance for the biosphere – including humankind. Aiming to describe them in as much detail as possible, several large-scale kinetic models have been constructed, for example to simulate marine ecosystems. Yet is there a way to circumvent the significant computational constraints and the widespread lack of kinetic data, while still arriving at predictions within the error margin of in-field measurements? To start investigating this question, the work here was performed with the plan of creating the simplest co-culture model conceivable and analyzing the effects of changes to the system variables and parameters. A kinetic model based on the Monod-equation was put together describing the syntrophic growth of an autotroph requiring anorganic carbon and producing organic carbon, and a heterotroph mineralizing the organic carbon back to inorganic. The composition of these two organismic fractions varies with the initial parameters. However, an eventual steady state with stable population numbers is unaffected by this and only requires a balance between growth and death rate. Death in particular is essential to ensure (organic) carbon redistribution to prevent the extinction of one of the co-culture members. Additionally, two metabolic models of Synechococcus elongatus PCC7942 and Escherichia coli K12 substr. MG1655 were used to perform dynamic flux balance analysis (dFBA). dFBA shows similar results with a stable steady state. Fitting of the dFBA results to the kinetic model gives further insights into some of the parameters, the yield factor in particular. Comparisons with literature are added to put the work in context and show that despite the simplistic nature of the model, the proof-of-concept results are valid enough to warrant further research into this direction.
Markus Schürmann: "Modelling glycolysis of human erythrocytes as a thermodynamic energy converter"
Based on the work of Rapoport and Heinrich, a model of glycolysis of human erythrocythes as a thermodynamic energy converter is presented in this work, which can be used both for investigations and as a basis for a future, more comprehensive model. It consists of a sim plified version of the Embden-Meyerhof-Parnas pathway, the Rapoport-Luebering shunt as a bypass of glycolysis in erythrocytes and the lactate dehydrogenase reaction, serving for redox and thermodynamic balance. In addition to kinetic, the model now also takes thermodynamic properties into account. Simulations over time show the system to reach steady state with convincing metabolite concentrations relative to the in vivo conditions. Based on this, the sys tem behaviour under varying demands of energy gets investigated, elucidating the role of the bypass as an energy buffer. Furthermore, the energy dissipation of the system gets calculated, determining the fractions of each reaction on the overall energy gradient and the efficiency as an energy converter under different conditions.
Marvin van Aalst: "Isotope-labelled modeling of the Calvin-Benson-Bassham cycle"
The Calvin-Benson-Bassham (CBB) cycle is used by many photosynthetic organisms to fix carbon dioxide. Thus it is one of the most important biochemical pathways on earth. Although it was proposed in 1954, the exact structure of the cycle is still under debate. The photosynthetic Gibbs effect, discovered via the asymmetrical incorporation of radioactive 14CO2 into hexoses, is widely accepted as being in accord with the proposed CBB reaction scheme, but the given arguments have been mostly qualitatively. In this work a mathematical model is used to provide a quantitative explanation of the photosynthetic Gibbs effect.
Ellen Oldenburg: "Mathematische Untersuchungen des bakteriellen Wachstums mit Ressourcenallokationsmodellen"
Viele biochemische Details des Bakterienmetabolismus und der Proteinsynthese wurden in den letzten 50 Jahren aufgeklärt, und es wurde gezeigt, dass auf molekularer Ebene Synthese, Abbau und Regulation über komplexe, miteinander verbundene Netzwerke durchgeführt werden. Dies geschieht durch Kinetiken, die nicht linear von den Konzentrationen der Reaktanten abhängen [5]. Dennoch erscheinen auf der physiologischen Ebene einfache empirische Beziehungen; diese werden als „Wachstumsgesetze“ bezeichnet [12]. Das Modell von Weiße et al. [13]wird verwendet, um das Verhalten des bakteriellen Wachstums unter bestimmten Bedingungen und Inhibierungen zu zeigen, von denen sich dann die bakteriellen Wachstumsgesetze ableiten lassen. Eine alternative Ableitung eines bakteriellen Wachstumsmodells wird auch von Scott et al. gegeben [11]. Die als "bakterielle Wachstumsgesetze"bezeichneten Phänomene sind empirische Beziehungen zwischen der Wachstumsrate und dem ribosomalen Massenanteil exponentiell wachsender Zellen [9] und dem Monod’schen Gesetz, das eine Michaelis-Menten-Beziehung zwischen Wachstumsrate und extrazellulärem Nährstoff feststellt [10]. In dieser Arbeit wird die Aussagekraft des Modells von Weiße et al. untersucht. Dieses Modell wird mit Daten belegt und die Ergebnisse mit denen aus der alternativen Ableitung von Scott et al. verglichen.
Saschas Ollertz: "A theoretical strudy of regulatory mechanisms in the chloroplast-located aspartate pathway"
Amino acids perform a variety of tasks in organisms, including being the building blocks of proteins. They are therefore important for every organism with its own metabolism. In Arabidopsis thaliana, some of these amino acids are synthesized in the chloroplast fromtheir common precursor aspartate. These are the essential amino acids lysine, threonine, isoleucine and methione. In order to investigate the regulation of their synthesis mathe-matically, the model of this pathway by Curien et al. (2009) was implemented, extended and investigated in this thesis. The key result of this work is that this model is extremely robust against external constraints and can compensate for them through its regulation.
Master's theses
Tobias Pfennig: Developing a computational model of photosynthesis in cyanobacteria
Cyanobacteria are ecologically relevant and are quickly gaining popularity in biotechnological applications. Their plant-like photosynthetic ability and shorter generation times make them excellent tools for light-driven production. However, cyanobacterial photosynthesis is still not fully understood. Much experimental effort has been put into elucidating the mechanisms of light capture and usage, but few theoretical analyses exist. This insufficiency deprives photosynthesis research of helpful support.
Furthermore, some published models are highly complicated and lack important cyanobacterial characteristics. Therefore, we developed a dynamic electron transport chain model of Synechocystis PCC 6803 addressing these concerns. Using up-to-date literature and simple kinetic descriptions, the model is designed to be representative and understandable. We show that the model describes the Electron Transport Chain's (ETC) internal redox changes well by reproducing experimental fluorescence measurements. Remarkably, we see that the light spectrum is critical to productivity and the internal redox state of electron transport. The enzyme-level resolution also allowed us to identify the high photosynthetic control of photosystem 2 and the cell's capability for flux control on the level of ferredoxin. This work demonstrates the importance of organism-specificity when modelling and hopes to support the research and usage of cyanobacteria.
Louis Goldschmidt: "Calorimetric Analysis of Saccharomyces cerivisiae Growth in a custom-built Bioreactor"
The heat produced by microbial culture systems is a quantity of interest, as it enables monitoring of metabolic activity and estimation of the thermodynamic efficiency of conversion of nutrients to biomass. Commercially available bioreactor systems with calorimetric capabilities are often expensive and designed with a specific task in mind, which makes the development of custom cultivation devices from available components a cheap and highly modular alternative. In this thesis, the design of a custom calorimetric bioreactor setup is presented. The device is calibrated based on the heat released during the solution of different amounts of sodium hydroxide. This is followed up by an evaluation of the setup over the course of a microbial cultivation experiment, throughout which the heat flow rates caused by the metabolic activity of the /Saccharomyces cerevisiae/ strain IFO 0233 are observed during batch and continuous cultivation. In the end, the mean heat dissipation per biomass carbon, a measure for the efficiency of biomass synthesis, is derived from the gathered data. The estimated value is then compared to literature values, which were determined during similar experiments, in order to evaluate the performance of the constructed system. The challenges associated with the operation of the setup as well as the simplifications and workarounds applied to mitigate them over the course of the evaluation are discussed in detail with regard to the quality of the produced data. In the end, even though the accuracy of the determined heat flow rates suffered due to inaccurate calibrations and uncertainty sources that were left unconsidered, the device was capable to produce a reasonable measurement of the metabolic heat released by the sustained microbial culture. Finally, I suggest several simple improvements that would make the utilization of the custom-built reactor system to study the metabolic dynamics of a microbial organism a viable option.
Thomas Wenske: "Association network of phage and prophage protein families"
The aim of this work was to create an association network between phages and prophages based on protein sequences. For this purpose 4,088 phage proteomes and 34,981 prophage proteomes were used, which were predicted from 18,872 prokaryotic genetic units (8,006 plasmids and 10,866 chromosomes). In addition, as further information, the connections between the prophages and their hosts (in the form of the genus) are present in the network. The network will serve to gain new knowledge about the similar phages and prophages, as well as distant relationships between phages, prophages and their hosts. For the calculation of the edge weighting in the network, local identities between the phage, the phage and the prophage and between the prophage proteins were first searched using BLAST, from which the orthologous hits were taken by searching reciprocal best hits. For the orthologous hits, the global identities were then calculated using a variation of needle. The program mcl then generated protein clusters. Subsequently, the final similarity between the different organism pairs was calculated using the jaccard index, which is calculated by the protein clusters, and the rooted average global identity of the protein pairs, and filtered by a cutoff of 10% similarity. This led to 196,354 similarities within the phages, 3,514,604 similarities within the prophages and 90,659 similarities between the phages and the prophages. There were proportionally more similar organism pairs between the phage than between the prophages and between the prophages and the phage and for the organism pairs between the phage and the prophage the similarity was on average the lowest. The generated network makes it possible to gain new insights into further protein similarity of more distant related organisms and thus further investigate evolutionary dynamics. However, it can also be observed that the data used represent only a fraction of the existing organisms in the environment, since many prophages could not be assigned to any known phage. In addition, it can be seen that for a small number of prokaryotic species there are more potential phages and predicted prophages in proportion, which are clustered together in the network due to their similarity. The network can be used as a tool to e.g. obtain information about distant organisms, which would not be found in a direct comparison between two organisms. Furthermore, it would also be possible to reconstruct new phages which have not yet been found, but which could be of great use in medicine or agriculture, for example, by means of the existing information of the prophages and if necessary also closely related phages.
Marvin van Aalst: "Optimality principles in leaf venation patterns"
Since leaves are the primary energy supplying organs of plants, their adaptation to the environment is one of the key steps a plant species has to take in order to vacate an ecological niche. It is thus not surprising, that plant leaves come in a variety of shapes and forms. But while evolutionary explanations to this variety are often given, the physico-chemical principles guiding it are largely unknown. This work focuses on one subset of this variety, the differences in vascular pattern formation. While a lot of work has been done on how the vascular pattern formation is established, surprisingly little work can be found on why these patterns were established in the first place. Plants have multiple competing objectives between which they have to choose in order to find a good adaptation to their environment. In this work I will focus on the first and second degree vein patterns, exploring the hypothesis, that leaf vascular architecture is a compromise between minimizing total vein length (minimizing investment) and minimizing transport distance
(maximizing efficiency)
Eric Behle: "Saccharification of Lignocellulose from Plant Cell Walls: a Theoretical Approach Based on Stochastic Simulations"
Plant cell walls from agricultural residues represent an abundant and unexploited raw material that intrinsically contains chemical energy in the form of polysaccharides. To face the need for new and renewable sources of energy this material is recently being considered as a promising alternative to conventional biofuel production processes. The lignocellulose structure of plant cell walls is composed of cellulose, hemicellulose and lignin, whose relative arrangement and abundance determine the mechanical properties of the plant tissue. The extraction of the simple fermentable sugars from the lignocellulose matrix is a complex chemical process that is time intensive and costly. A large fraction of the ongoing research in that field is led by industrial interest, and as a consequence the fundamental understanding of this complex biochemical and biophysical process is yet to be unraveled. Towards this objective, during this project, a modeling approach based on stochastic numerical simulations was developed to investigate the saccharification dynamics of a lignocellulose microfibril. The spatial arrangement and composition of the microfibril polymers were represented in silico, and, using a Gillespie algorithm, the saccharification process resulting from the action of a set of well studied digesting enzymes was performed. To calibrate the kinetic parameters of the model, substrate composition and saccharification dynamics data from literature were utilized. From this it was examined, how the internal structure of the substrate impacts the availability of the sugars, and how the mechanistic properties of the individual enzymes determine their synergism and drive the saccharification process.