Durch die Einführung eines neuen fallpauschalierten Abrechnungssystems (Diangnosis related groups, DRG) sowie zahlreiche neue Maßnahmen der Qualitätssicherung wächst der Anteil administrativer Tätigkeiten für die klinisch tätigen Ärztinnen und Ärzte. Diese Entwicklung hat nicht zuletzt auch Auswirkungen auf die zeitlichen Ressourcen, die für die ärztlichen Kernaufgaben zur Verfügung stehen. In Bereichen mit intensiver klinischer Forschung steigen darüberhinaus auch die Anforderungen an die forschungsbegleitende Dokumentation. Eine effiziente Unterstützung bei der Kalkulation, Planung und Durchführung von Therapien durch Informationstechnologische (IT-) Systeme kann eine sinnvolle Entlastung der Ärzte darstellen. Es wurde ein einsatzfähiger Prototyp entwickelt, prgrammiert und implementiert. Anschließend wurde das Ergebnis mit potentiellen Anwendern der Software evaluiert. / With the introduction of diagnoses related groups (DRG) and with other task for quality control, the part of adminitrative task in medical work rises every day. This development has an impact on the timely resources that are available for medical core issues. In departments with intense clinical research, the requirements for research-relatet documentation are rising. An effectiv support in calculation, planing and performing of therapies by modern information technology can bei a vital factor for clincial works. A ready-to-use prototyp of a computer-aided chemotherapy-system has been develepoed, programmed and implemented. Afterwards, the results were evalueted by possible users of this system.

The aim of this study is to illustrate the state of-the art of technical tools which allow the user to build the lexicon of a Swahili text. Different kinds of statistical information can also be extracted from the text with the aid of tailor made software. The basic operation in building the lexicon of a text is lemmatization, i. e extracting the lemma from the forms contained in the text. Once the lemma list is ready it can be converted into a list of entties, to be filled according to selected criteria.

Adrenal steroid hormones, which regulate a plethora of physiological functions, are produced via tightly controlled pathways. Adrenal hormone excess associates with clinical conditions impacting metabolism and cardiovascular and immune function. Aldosterone and cortisol producing enzymes, CYP11B2 and CYP11B1, share 93% homology requiring highly selective drugs for pharmacological treatment. Investigations of these pathways, based on experimental data, can be facilitated by computational modeling for calculations of metabolic rate alterations. Such systems can be utilized in a variety of applications including investigating effects of endocrine-disrupting chemicals, drugs, gene manipulations. On a human level, regression modelling involving use of clinical data can contribute in defining effects of such diseases or providing reference data for characterizing normal values of physiological processes. The main subject of this thesis was the development of modeling techniques that would benefit basic and clinical research by supplying means for investigating adrenal related dysfunctions and disease. As a first approach, we used a model system, based on mass balance and mass reaction equations, to kinetically evaluate adrenal steroidogenesis in human adrenal cortex-derived NCI H295R cells. For this purpose a panel of 10 steroids was measured by liquid chromatographic-tandem mass spectrometry. Time-dependent changes in cell incubate concentrations of steroids were measured after incubation with angII, forskolin and abiraterone. Model parameters were estimated based on experimental data using weighted least square fitting. Time-dependent angII- and forskolin-induced changes were observed for incubate concentrations of precursor steroids with peaks that preceded maximal increases in aldosterone and cortisol. Inhibition of 17-alpha-hydroxylase/17,20-lyase with abiraterone resulted in increases in upstream precursor steroids and decreases in downstream products. Derived model parameters, including rate constants of enzymatic processes, appropriately quantified observed and expected changes in metabolic pathways at multiple conversion steps. Our data from our first approach demonstrated limitations of single time point measurements and the importance of assessing pathway dynamics in studies of adrenal cortical cell line steroidogenesis. Despite the benefits from a computational approach in comparison to the single time point studies, this kind of modeling demonstrated certain limitations regarding effort, reproducibility and costs. Therefor a second study was conducted in order to address such limitations. As a second approach we introduced an effective in vitro assay for evaluation of steroidogenic enzyme kinetics based on intracellular flux calculations. H295RA cells were cultured in chambers (µ-Slide, Ibidi) under constant medium flow. Four hourly samples were collected (control samples), followed by collections over an additional four hours after treatment with either fadrozole (10nM), metyrapone (10uM), ASI_191 (5nM), a novel CYP11B2 inhibitor or ASI_254 (100nM), a newly synthesized CYP17 inhibitor. Mass spectrometric measurements of multiple steroids combined with linear system computational modeling facilitated calculation of intracellular flux rates at different steroidogenic pathway steps and assessment of the selectivity of drugs for those specific steps. While treatment with fadrozole, metyrapone and ASI_191 all resulted in reductions in fluxes of aldosterone, corticosterone and cortisol production, treatment with ASI_254 led to increased flux through the mineralocorticoid pathway and increased production of aldosterone with reduced production of steroids downstream of CYP17. Comparisons of changes in intracellular fluxes revealed much higher selectivity of ASI_191 for CYP11B2 over CYP11B1 compared to fadrozole or metyrapone. Our study demonstrates the advantages of continuous culture systems over static systems for studying effects of steroidogenic inhibitors. By culturing cells under perfusion the methodology establishes a more realistic model for investigating drug effects, provides for simple and rapid calculations of intracellular fluxes and offers a robust method for drug screening or in vitro investigations of metabolic mechanisms. As a third approach we utilized LC-MS/MS derived plasma concentrations for each of 525 normotensive and hypertensive volunteers with (n=227) and without (n=298) hypertension in combination with regression modeling for the extraction of age and gender-adjusted reference intervals. Values of 8 steroids (pregnenolone, 11-deoxycorticosterone, corticosterone, 17-hydroxyprogesterone, cortisone, dehydroepiandrosterone, dehydroepiandrosterone-sulfate, androstenedione) versus age and gender were modelled via multivariate fractional polynomial analysis successfully providing with 0.5 and 99.5% reference intervals as a function of age and gender.:Contents Acknowledgements 4 Publication Note 5 Contents 6 Abbreviations 9 1. Introduction 10 1.1. The adrenal gland 10 1.1.1. Adrenal cortex steroids regulation 10 1.1.2. Physiological metabolic processes of steroidogenesis 13 1.2. Adrenal cortical dysfunction and therapeutic challenges 15 1.3. The adrenocortical cell line NCI –H295R 16 1.4. Chemical regulators of adrenal steroidogenesis 17 1.5. Computational mechanistic modelling of adrenal metabolism 18 1.5.1. Static cell culture models for characterizing pathway dynamics 18 1.5.2. Steady state as a tool for intracellular fluxes estimation 19 2. Hypothesis and aims 20 3. Materials and methods 22 3.1. Experimental overview 22 3.1.1. Computational mechanistic modeling of steroid metabolism 22 3.1.2 A steady state system for in vitro evaluation of steroidogenic pathway dynamics 24 3.1.3 Calculation of age and gender adjusted reference intervals for plasma adrenal steroids for healthy population 27 3.2 Liquid chromatography – tandem mass spectrometry 29 3.3. Static cell culture mechanistic model 29 3.3.1. Cell culture conditions 29 3.3.2. Computational model representation for steroid metabolism and cell proliferation using ODE systems 29 3.3.3. Metabolic pathways 30 3.3.4. Transport processes 32 3.3.5. Parameter estimation 34 3.4. Steady state model system for computation of pathway kinetics 35 3.4.1. Cell culture conditions for continuous flow culture system 35 3.4.2 Steroid secretion rates and intracellular fluxes calculation under steady state conditions 37 3.4.3 Statistical analysis 40 3.5. Regression and classification models for diagnostic clinical studies 40 3.5.1. Age and sex adjusted reference intervals for adrenal steroids in plasma – patient data collection 40 3.5.2. Mathematical description of multivariate fractional polynomial analysis 41 4. Results 43 4.1. Computational analysis of steroid profiling in NCI H295R cells 43 4.1.1. Transport and metabolic pathway modeling 43 4.1.2. Angiotensin II and forskolin stimulation 43 4.1.3. Abiraterone treatment 46 4.2. Steady state model for in vitro evaluation of steroidogenic pathway dynamics 48 4.2.1. Continuous flow culture steroid profiling 48 4.2.2. Secretion rates, intracellular flux rates and relative changes in rate constants 51 4.2.3. Cell number and viability 53 4.3. Reference intervals for adrenal steroid plasma concentrations 55 5. Discussion 61 Appendix A – Static culture model 70 A1. ODE system equations of the static culture model 70 A2. Jsim MML example code of the ODE system implementation of the static culture model 74 A3. Supplementary data for static culture model derived data 79 Appendix B – Steady state model 84 B1. Mathematical derivation of analytical solution for intracellular flux calculation of the steady state model 84 B2. Equations describing steroid production in cells and flowing medium of the steady state model 86 B3. Quadratic programming formulation and solution of upstream steroidogenic pathways system of the steady state model 91 B4. Calculation of reaction rate constant relative changes 92 B5. Python implementation of secretion rates, flux rates and rate constant relative changes calculation 94 B6. Numerical example of intracellular flux calculation. 98 B7. Supplementary material for steady state model derived data 103 Appendix C – Age and gender-adjusted reference intervals 110 Zusammenfasung 112 Summary 113 Literature 116

The aim of this study is to illustrate the state of-the art of technical tools which allow the user to build the lexicon of a Swahili text. Different kinds of statistical information can also be extracted from the text with the aid of tailor made software. The basic operation in building the lexicon of a text is lemmatization, i. e extracting the lemma from the forms contained in the text. Once the lemma list is ready it can be converted into a list of entties, to be filled according to selected criteria.

Background: The clinical outcome of hip resurfacing (HR) as a demanding surgical technique associated with a substantial learning curve depends on the position of the femoral component. The aim of the study was to investigate the effects of the level of surgical experience on computer-assisted imageless navigation concerning precision of femoral component positioning, notching, and oversizing rate, as well as operative time. Methods: Three surgeons with different levels of experience in both HR and computer-assisted surgery (CAS) prepared the femoral heads of 54 synthetic femurs using the Durom TM Hip Resurfacing (Zimmer, Warsaw, IN, USA) system. Each surgeon prepared a total of 18 proximal femurs using the Navitrack® system (ORTHOsoft Inc., Montreal, Canada) or the conventional free-hand Durom TM K-wire positioning jig. The differences between planned and postoperative stem shaft angle (SSA) and anteversion angle in standardized x-rays were measured and the operative time, not including the time for calibrating the CAS-system, was documented. Notching was evaluated by the three surgeons in a randomized manner. Oversizing was determined by the difference of the preoperative determined cap and the cap size advised by the CAS-system. Results: CAS significantly reduced the overall mean deviation between planned and postoperative SSA in comparison with the conventional procedure (mean ± SD, 1 ± 1.7° vs. 7.4 ± 4.4°, P <0.01) regardless of the surgeon’s level of experience. The incidence of either varus or valgus SSA deviations exceeding 5° were 1/27 for CAS and 15/27 for the conventional method, respectively (P<0.001), corresponding to a reduction by 97%. Using CAS, the rate of notching was reduced by 100%. Conclusions: The accuracy of femoral HR component orientation is significantly increased by use of CAS regardless of the surgeon’s level of experience in our preclinical study. Thus, imageless computer-assisted navigation can be a valuable tool to improve implant positioning in HR for surgeons at any stage of their learning curve.

During tissue invasion individual tumor cells exhibit two interconvertible migration modes, namely mesenchymal and amoeboid migration. The cellular microenvironment triggers the switch between both modes, thereby allowing adaptation to dynamic conditions. It is, however, unclear if this amoeboid-mesenchymal migration plasticity contributes to a more effective tumor invasion. We address this question with a mathematical model, where the amoeboid-mesenchymal migration plasticity is regulated in response to local extracellular matrix resistance. Our numerical analysis reveals that extracellular matrix structure and presence of a chemotactic gradient are key determinants of the model behavior. Only in complex microenvironments, if the extracellular matrix is highly heterogeneous and a chemotactic gradient directs migration, the amoeboid-mesenchymal migration plasticity allows a more widespread invasion compared to the non-switching amoeboid and mesenchymal modes. Importantly, these specific conditions are characteristic for in vivo tumor invasion. Thus, our study suggests that in vitro systems aiming at unraveling the underlying molecular mechanisms of tumor invasion should take into account the complexity of the microenvironment by considering the combined effects of structural heterogeneities and chemical gradients on cell migration.

In dieser Arbeit werden Fortschritte in der Elektronenptychographie vorgestellt, die ihre Vielseitigkeit als Technik in der Elektronen-Phasenkontrastmikroskopie verbessern. Anstatt sich auf eine hochauflösende Elektronenoptik zu stützen, rekonstruiert die Ptychographie die Proben auf der Grundlage ihrer kohärenten Beugungssignale mit Hilfe von Berechnungsalgorithmen. Dieser Ansatz ermöglicht es, die Grenzen der konventionellen, auf Optik basierenden Elektronenmikroskopie zu überwinden und eine noch nie dagewesene sub-Angstrom Auflösung in den resultierenden Bildern zu erreichen. In dieser Arbeit werden zunächst die theoretischen, experimentellen und algorithmischen Grundlagen der Elektronenptychographie vorgestellt und in den Kontext der bestehenden rastergestützten Elektronenmikroskopietechniken gestellt. Darüber hinaus wird ein alternativer ptychographischer Phasengewinnungsalgorithmus entwickelt und seine Leistungsfähigkeit sowie die Qualität und räumliche Auflösung der Rekonstruktionen analysiert. Weiterhin befasst sich die Arbeit mit der Integration von Methoden des maschinellen Lernens in die Elektronenptychographie und schlägt einen spezifischen Ansatz zur Verbesserung der Rekonstruktionsqualität unter suboptimalen Versuchsbedingungen vor. Außerdem wird die Kombination von Ptychographie mit Defokusserienmessungen hervorgehoben, die eine verbesserte Tiefenauflösung bei ptychographischen Rekonstruktionen ermöglicht und uns somit dem ultimativen Ziel näher bringt, quantitative Rekonstruktionen von beliebig dicker Proben mit atomarer Auflösung in drei Dimensionen zu erzeugen. Der letzte Teil der Arbeit stellt einen Paradigmenwechsel bei den Scananforderungen für die Ptychographie vor und zeigt Anwendungen dieses neuen Ansatzes unter Bedingungen niedriger Dosis. / This thesis presents advancements in electron ptychography, enhancing its versatility as an electron phase-contrast microscopy technique. Rather than relying on high-resolution electron optics, ptychography reconstructs specimens based on their coherent diffraction signals using computational algorithms. This approach allows us to surpass the limitations of conventional optics-based electron microscopy, achieving an unprecedented sub-Angstrom resolution in the resulting images. The thesis initially introduces the theoretical, experimental, and algorithmic principles of electron ptychography, contextualizing them within the landscape of existing scanning-based electron microscopy techniques. Additionally, it develops an alternative ptychographic phase retrieval algorithm, analyzing its performance and also the quality and the spatial resolution of its reconstructions. Moreover, the thesis delves into the integration of machine learning methods into electron ptychography, proposing a specific approach to enhance reconstruction quality under suboptimal experimental conditions. Furthermore, it highlights the fusion of ptychography with defocus series measurements, offering improved depth resolution in ptychographic reconstructions, which therefore brings us closer to the ultimate goal of quantitative reconstructions of arbitrarily thick specimens at atomic resolution in three dimensions. The final part of the thesis introduces a paradigm shift in scanning requirements for ptychography and showcases applications of this novel approach under low-dose conditions.

Background: The clinical outcome of hip resurfacing (HR) as a demanding surgical technique associated with a substantial learning curve depends on the position of the femoral component. The aim of the study was to investigate the effects of the level of surgical experience on computer-assisted imageless navigation concerning precision of femoral component positioning, notching, and oversizing rate, as well as operative time. Methods: Three surgeons with different levels of experience in both HR and computer-assisted surgery (CAS) prepared the femoral heads of 54 synthetic femurs using the Durom TM Hip Resurfacing (Zimmer, Warsaw, IN, USA) system. Each surgeon prepared a total of 18 proximal femurs using the Navitrack® system (ORTHOsoft Inc., Montreal, Canada) or the conventional free-hand Durom TM K-wire positioning jig. The differences between planned and postoperative stem shaft angle (SSA) and anteversion angle in standardized x-rays were measured and the operative time, not including the time for calibrating the CAS-system, was documented. Notching was evaluated by the three surgeons in a randomized manner. Oversizing was determined by the difference of the preoperative determined cap and the cap size advised by the CAS-system. Results: CAS significantly reduced the overall mean deviation between planned and postoperative SSA in comparison with the conventional procedure (mean ± SD, 1 ± 1.7° vs. 7.4 ± 4.4°, P <0.01) regardless of the surgeon’s level of experience. The incidence of either varus or valgus SSA deviations exceeding 5° were 1/27 for CAS and 15/27 for the conventional method, respectively (P<0.001), corresponding to a reduction by 97%. Using CAS, the rate of notching was reduced by 100%. Conclusions: The accuracy of femoral HR component orientation is significantly increased by use of CAS regardless of the surgeon’s level of experience in our preclinical study. Thus, imageless computer-assisted navigation can be a valuable tool to improve implant positioning in HR for surgeons at any stage of their learning curve.

During tissue invasion individual tumor cells exhibit two interconvertible migration modes, namely mesenchymal and amoeboid migration. The cellular microenvironment triggers the switch between both modes, thereby allowing adaptation to dynamic conditions. It is, however, unclear if this amoeboid-mesenchymal migration plasticity contributes to a more effective tumor invasion. We address this question with a mathematical model, where the amoeboid-mesenchymal migration plasticity is regulated in response to local extracellular matrix resistance. Our numerical analysis reveals that extracellular matrix structure and presence of a chemotactic gradient are key determinants of the model behavior. Only in complex microenvironments, if the extracellular matrix is highly heterogeneous and a chemotactic gradient directs migration, the amoeboid-mesenchymal migration plasticity allows a more widespread invasion compared to the non-switching amoeboid and mesenchymal modes. Importantly, these specific conditions are characteristic for in vivo tumor invasion. Thus, our study suggests that in vitro systems aiming at unraveling the underlying molecular mechanisms of tumor invasion should take into account the complexity of the microenvironment by considering the combined effects of structural heterogeneities and chemical gradients on cell migration.

Cancer is a very complex genetic disease driven by combinations of mutated genes. This complexity strongly complicates the identification of driver genes and puts enormous challenges to reveal how they influence cancerogenesis, prognosis or therapy response. Thousands of molecular profiles of the major human types of cancer have been measured over the last years. Apart from well-studied frequently mutated genes, still only little is known about the role of rarely mutated genes in cancer or the interplay of mutated genes in individual cancers. Gene expression and mutation profiles can be measured routinely, but computational methods for the identification of driver candidates along with the prediction of their potential impacts on downstream targets and clinically relevant characteristics only rarely exist. Instead of only focusing on frequently mutated genes, each cancer patient should better be analyzed by using the full information in its cancer-specific molecular profiles to improve the understanding of cancerogenesis and to more precisely predict prognosis and therapy response of individual patients. This requires novel computational methods for the integrative analysis of molecular cancer data. A promising way to realize this is to consider cancer as a disease of cellular networks. Therefore, I have developed a novel network-based approach for the integrative analysis of molecular cancer data over the last years. This approach directly learns gene regulatory networks form gene expression and copy number data and further enables to quantify impacts of altered genes on clinically relevant downstream targets using network propagation. This habilitation thesis summarizes the results of seven of my publications. All publications have a focus on the integrative analysis of molecular cancer data with an overarching connection to the newly developed network-based approach. In the first three publications, networks were learned to identify major regulators that distinguish characteristic gene expression signatures with applications to astrocytomas, oligodendrogliomas, and acute myeloid leukemia. Next, the central publication of this habilitation thesis, which combines network inference with network propagation, is introduced. The great value of this approach is demonstrated by quantifying potential direct and indirect impacts of rare and frequent gene copy number alterations on patient survival. Further, the publication of the corresponding user-friendly R package regNet is introduced. Finally, two additional publications that also strongly highlight the value of the developed network-based approach are presented with the aims to predict cancer gene candidates within the region of the 1p/19q co-deletion of oligodendrogliomas and to determine driver candidates associated with radioresistance and relapse of prostate cancer. All seven publications are embedded into a brief introduction that motivates the scientific background and the major objectives of this thesis. The background is briefly going from the hallmarks of cancer over the complexity of cancer genomes down to the importance of networks in cancer. This includes a short introduction of the mathematical concepts that underlie the developed network inference and network propagation algorithms. Further, I briefly motivate and summarize my studies before the original publications are presented. The habilitation thesis is completed with a general discussion of the major results with a specific focus on the utilized network-based data analysis strategies. Major biologically and clinically relevant findings of each publication are also briefly summarized.