Hexokinase 2 (Hxk2) of Saccharomyces cerevisiae is a dual function hexokinase, acting as a glycolytic enzyme and being involved in the transcriptional regulation of glucose-repressible genes. Relief from glucose repression is accompanied by phosphorylation of Hxk2 at serine 15, which has been attributed to the protein kinase Tda1. To explore the role of Tda1 beyond Hxk2 phosphorylation, the proteomic consequences of TDA1 deficiency were investigated by difference gel electrophoresis (2D-DIGE) comparing a wild type and a Δtda1 deletion mutant. To additionally address possible consequences of glucose repression/derepression, both were grown at 2% and 0.1% (w/v) glucose. A total of eight protein spots exhibiting a minimum twofold enhanced or reduced fluorescence upon TDA1 deficiency was detected and identified by mass spectrometry. Among the spot identities are—besides the expected Hxk2—two proteoforms of hexokinase 1 (Hxk1). Targeted proteomics analyses in conjunction with 2D-DIGE demonstrated that TDA1 is indispensable for Hxk2 and Hxk1 phosphorylation at serine 15. Thirty-six glucose-concentration-dependent protein spots were identified. A simple method to improve spot quantification, approximating spots as rotationally symmetric solids, is presented along with new data on the quantities of Hxk1 and Hxk2 and their serine 15 phosphorylated forms at high and low glucose growth conditions. The Δtda1 deletion mutant exhibited no altered growth under high or low glucose conditions or on alternative carbon sources. Also, invertase activity, serving as a reporter for glucose derepression, was not significantly altered. Instead, an involvement of Tda1 in oxidative stress response is suggested.

In a world with rapidly increasing population that competes for the earth’s limited resources, cooperation is crucial. While research showed that empathizing with another individual in need enhances prosociality, it remains unclear whether correctly inferring the other’s inner, mental states on a more cognitive level (i.e., mentalizing) elicits helping behavior as well. We applied a video-based laboratory task probing empathy and a performance measure of mentalizing in adult volunteers (N = 94) and assessed to which extent they were willing to help the narrators in the videos. We replicate findings that an empathy induction leads to more prosocial decisions. Crucially, we also found that correct mentalizing increases the willingness to help. This evidence helps clarify an inconsistent picture of the relation between mentalizing and prosociality.

Acute Respiratory Distress Syndrome (ARDS) is common in COVID-19 patients and is associated with high mortality. The aim of this observational study was to describe patients’ characteristics and outcome, identifying potential risk factors for in-hospital mortality and for developing Long-COVID symptoms. This retrospective study included all patients with COVID-19 associated ARDS (cARDS) in the period from March 2020 to March 2021 who were invasively ventilated at the intensive care unit (ICU) of the University Hospital Dresden, Germany. Between October 2021 and December 2021 patients discharged alive (at minimum 6 months after hospital discharge—midterm survival) were contacted and interviewed about persistent symptoms possibly associated with COVID-19 as well as the quality of their lives using the EQ-5D-5L-questionnaire. Long-COVID was defined as the occurrence of one of the symptoms at least 6 months after discharge. Risk factors for mortality were assessed with Cox regression models and risk factors for developing Long-COVID symptoms by using relative risk (RR) regression. 184 Patients were included in this study (male: n = 134 (73%), median age 67 (range 25–92). All patients were diagnosed with ARDS according to the Berlin Definition. 89% of patients (n = 164) had severe ARDS (Horovitz-index < 100 mmHg). In 27% (n = 49) extracorporeal membrane oxygenation was necessary to maintain gas exchange. The median length of in-hospital stay was 19 days (range 1–60). ICU mortality was 51%, hospital mortality 59%. Midterm survival (median 11 months) was 83% (n = 55) and 78% (n = 43) of these patients presented Long-COVID symptoms with fatigue as the most common symptom (70%). Extreme obesity (BMI > 40 kg/m2) was the strongest predictor for in-hospital mortality (hazard ratio: 3.147, confidence interval 1.000–9.897) and for developing Long-COVID symptoms (RR 1.61, confidence interval 1.26–2.06). In-hospital mortality in severe cARDS patients was high, but > 80% of patients discharged alive survived the midterm observation period. Nonetheless, most patients developed Long-COVID symptoms. Extreme obesity with BMI > 40 kg/m2 was identified as independent risk factor for in-hospital mortality and for developing Long-COVID symptoms.

Human temporal bone specimens are used in experiments measuring the sound transfer of the middle ear, which is the standard method used in the development of active and passive middle ear implants. Statistical analyses of these experiments usually require that the TB samples are representative of the population of non-pathological middle ears. Specifically, this means that the specimens must be mechanically well-characterized. We present an in-depth statistical analysis of 478 data sets of middle ear transfer functions (METFs) from different laboratories. The data sets are preprocessed and various contributions to the variance of the data are evaluated. We then derive a statistical range as a reference against which individual METF measurements may be validated. The range is calculated as the two-sided 95% tolerance interval at audiological frequencies. In addition, the mean and 95% confidence interval of the mean are given as references for assessing the validity of a sample group. Finally, we provide a suggested procedure for measuring METFs using the methods described herein.

Adipose tissue expansion involves both differentiation of new precursors and size increase of mature adipocytes. While the two processes are well balanced in healthy tissues, obesity and diabetes type II are associated with abnormally enlarged adipocytes and excess lipid accumulation. Previous studies suggested a link between cell stiffness, volume and stem cell differentiation, although in the context of preadipocytes, there have been contradictory results regarding stiffness changes with differentiation. Thus, we set out to quantitatively monitor adipocyte shape and size changes with differentiation and lipid accumulation. We quantified by optical diffraction tomography that differentiating preadipocytes increased their volumes drastically. Atomic force microscopy (AFM)-indentation and -microrheology revealed that during the early phase of differentiation, human preadipocytes became more compliant and more fluid-like, concomitant with ROCK-mediated F-actin remodelling. Adipocytes that had accumulated large lipid droplets were more compliant, and further promoting lipid accumulation led to an even more compliant phenotype. In line with that, high fat diet-induced obesity was associated with more compliant adipose tissue compared to lean animals, both for drosophila fat bodies and murine gonadal adipose tissue. In contrast, adipose tissue of diabetic mice became significantly stiffer as shown not only by AFM but also magnetic resonance elastography. Altogether, we dissect relative contributions of the cytoskeleton and lipid droplets to cell and tissue mechanical changes across different functional states, such as differentiation, nutritional state and disease. Our work therefore sets the basis for future explorations on how tissue mechanical changes influence the behaviour of mechanosensitive tissue-resident cells in metabolic disorders.

Despite advances in nuclease-based genome editing technologies, correcting human disease-causing genomic inversions remains a challenge. Here, we describe the potential use of a recombinase-based system to correct the 140 kb inversion of the F8 gene frequently found in patients diagnosed with severe Hemophilia A. Employing substrate-linked directed molecular evolution, we develop a coupled heterodimeric recombinase system (RecF8) achieving 30% inversion of the target sequence in human tissue culture cells. Transient RecF8 treatment of endothelial cells, differentiated from patient-derived induced pluripotent stem cells (iPSCs) of a hemophilic donor, results in 12% correction of the inversion and restores Factor VIII mRNA expression. In this work, we present designer-recombinases as an efficient and specific means towards treatment of monogenic diseases caused by large gene inversions.

The taxonomy of the Mediterranean Aristolochia pallida complex has been under debate since several decades with the following species currently recognized: A. pallida, A. lutea, A. nardiana, A. microstoma, A. merxmuelleri, A. croatica, and A. castellana. These taxa are distributed from Iberia to Turkey. To reconstruct phylogenetic and biogeographic patterns, we employed cpDNA sequence variation using both noncoding (intron and spacer) and protein-coding regions (i.e., trnK intron, matK gene, and trnK-psbA spacer). Our results show that the morphology-based traditional taxonomy was not corroborated by our phylogenetic analyses. Aristolochia pallida, A. lutea, A. nardiana, and A. microstoma were not monophyletic. Instead, strong geographic signals were detected. Two major clades, one exclusively occurring in Greece and a second one of pan-Mediterranean distribution, were found. Several subclades distributed in Greece, NW Turkey, Italy, as well as amphi-Adriatic subclades, and a subgroup of southern France and Spain, were revealed. The distribution areas of these groups are in close vicinity to hypothesized glacial refugia areas in the Mediterranean. According to molecular clock analyses the diversification of this complex started around 3–3.3 my, before the onset of glaciation cycles, and the further evolution of and within major lineages falls into the Pleistocene. Based on these data, we conclude that the Aristolochia pallida alliance survived in different Mediterranean refugia rarely with low, but often with a high potential for range extension, and a high degree of morphological diversity.

Radiomics analyses commonly apply imaging features of different complexity for the prediction of the endpoint of interest. However, the prognostic value of each feature class is generally unclear. Furthermore, many radiomics models lack independent external validation that is decisive for their clinical application. Therefore, in this manuscript we present two complementary studies. In our modelling study, we developed and validated different radiomics signatures for outcome prediction after neoadjuvant chemoradiotherapy (nCRT) in patients with locally advanced rectal cancer (LARC) based on computed tomography (CT) and T2-weighted (T2w) magnetic resonance (MR) imaging datasets of 4 independent institutions (training: 122, validation 68 patients). We compared different feature classes extracted from the gross tumour volume for the prognosis of tumour response and freedom from distant metastases (FFDM): morphological and first order (MFO) features, second order texture (SOT) features, and Laplacian of Gaussian (LoG) transformed intensity features. Analyses were performed for CT and MRI separately and combined. Model performance was assessed by the area under the curve (AUC) and the concordance index (CI) for tumour response and FFDM, respectively. Overall, intensity features of LoG transformed CT and MR imaging combined with clinical T stage (cT) showed the best performance for tumour response prediction, while SOT features showed good performance for FFDM in independent validation (AUC = 0.70, CI = 0.69). In our external validation study, we aimed to validate previously published radiomics signatures on our multicentre cohort. We identified relevant publications on comparable patient datasets through a literature search and applied the reported radiomics models to our dataset. Only one of the identified studies could be validated, indicating an overall lack of reproducibility and the need of further standardization of radiomics before clinical application.

To evaluate the suitability of volume index measurement (VI) by either ultrasound (US) or computed tomography (CT) for the assessment of liver volume. Fifty-nine patients, 21 women, with a mean age of 66.8 ± 12.6 years underwent US of the liver followed immediately by abdominal CT. In US and CT imaging dorsoventral, mediolateral and craniocaudal liver diameters in their maximum extensions were assessed by two observers. VI was calculated by multiplication of the diameters divided by a constant (3.6). The liver volume determined by a manual segmentation in CT (“true liver volume”) served as gold standard. True liver volume and calculated VI determined by US and CT were compared using Bland–Altman analysis. Mean differences of VI between observers were − 34.7% (− 90.1%; 20.7%) for the US-based and 1.1% (− 16.1%; 18.2%) for the CT-based technique, respectively. Liver volumes determined by semi-automated segmentation, US-based VI and CT-based VI, were as follows: 1.500 ± 347cm3; 863 ± 371cm3; 1.509 ± 432cm3. Results showed a great discrepancy between US-based VI and true liver volume with a mean bias of 58.3 ± 66.9%, and high agreement between CT-based VI and true liver volume with a low mean difference of 4.4 ± 28.3%. Volume index based on CT diameters is a reliable, fast and simple approach for estimating liver volume and can therefore be recommended for clinical practice. The usage of US-based volume index for assessment of liver volume should not be used due to its low accuracy of US in measurement of liver diameters.

This thesis is divided into two parts dealing with the optimization problems in Markov decision processes (MDPs) and different neural network-based numerical solvers for partial differential equations (PDEs). In Part I we analyze the optimization problem arising in (partially observable) Markov decision processes using tools from algebraic statistics and information geometry, which can be viewed as neighboring fields of applied algebra and differential geometry, respectively. Here, we focus on infinite horizon problems and memoryless stochastic policies. Markov decision processes provide a mathematical framework for sequential decision-making on which most current reinforcement learning algorithms are built. They formalize the task of optimally controlling the state of a system through appropriate actions. For fully observable problems, the action can be selected knowing the current state of the system. This case has been studied extensively and optimizing the action selection is known to be equivalent to solving a linear program over the (generalized) stationary distributions of the Markov decision process, which are also referred to as state-action frequencies. In Chapter 3, we study partially observable problems where an action must be chosen based solely on an observation of the current state, which might not fully reveal the underlying state. We characterize the feasible state-action frequencies of partially observable Markov decision processes by polynomial inequalities. In particular, the optimization problem in partially observable MDPs is described as a polynomially constrained linear objective program that generalizes the (dual) linear programming formulation of fully observable problems. We use this to study the combinatorial and algebraic complexity of this optimization problem and to upper bound the number of critical points over the individual boundary components of the feasible set. Furthermore, we show that our polynomial programming formulation can be used to effectively solve partially observable MDPs using interior point methods, numerical algebraic techniques, and convex relaxations. Gradient-based methods, including variants of natural gradient methods, have gained tremendous attention in the theoretical reinforcement learning community, where they are commonly referred to as (natural) policy gradient methods. In Chapter 4, we provide a unified treatment of a variety of natural policy gradient methods for fully observable problems by studying their state-action frequencies from the standpoint of information geometry. For a variety of NPGs and reward functions, we show that the trajectories in state-action space are solutions of gradient flows with respect to Hessian geometries, based on which we obtain global convergence guarantees and convergence rates. In particular, we show linear convergence for unregularized and regularized NPG flows with the metrics proposed by Morimura and co-authors and Kakade by observing that these arise from the Hessian geometries of the entropy and conditional entropy, respectively. Further, we obtain sublinear convergence rates for Hessian geometries arising from other convex functions like log-barriers. We provide experimental evidence indicating that our predicted rates are essentially tight. Finally, we interpret the discrete-time NPG methods with regularized rewards as inexact Newton methods if the NPG is defined with respect to the Hessian geometry of the regularizer. This yields local quadratic convergence rates of these methods for step size equal to the inverse penalization strength, which recovers existing results as special cases. Part II addresses neural network-based PDE solvers that have recently experienced tremendous growth in popularity and attention in the scientific machine learning community. We focus on two approaches that represent the approximation of a solution of a PDE as the minimization over the parameters of a neural network: the deep Ritz method and physically informed neural networks. In Chapter 5, we study the theoretical properties of the boundary penalty for these methods and obtain a uniform convergence result for the deep Ritz method for a large class of potentially nonlinear problems. For linear PDEs, we estimate the error of the deep Ritz method in terms of the optimization error, the approximation capabilities of the neural network, and the strength of the penalty. This reveals a trade-off in the choice of the penalization strength, where too little penalization allows large boundary values, and too strong penalization leads to a poor solution of the PDE inside the domain. For physics-informed networks, we show that when working with neural networks that have zero boundary values also the second derivatives of the solution are approximated whereas otherwise only lower-order derivatives are approximated. In Chapter 6, we propose energy natural gradient descent, a natural gradient method with respect to second-order information in the function space, as an optimization algorithm for physics-informed neural networks and the deep Ritz method. We show that this method, which can be interpreted as a generalized Gauss-Newton method, mimics Newton’s method in function space except for an orthogonal projection onto the tangent space of the model. We show that for a variety of PDEs, natural energy gradients converge rapidly and approximations to the solution of the PDE are several orders of magnitude more accurate than gradient descent, Adam and Newton’s methods, even when these methods are given more computational time.:Chapter 1. Introduction 1 1.1 Notation and conventions 7 Part I. Geometry of Markov decision processes 11 Chapter 2. Background on Markov decision processes 12 2.1 State-action frequencies 19 2.2 The advantage function and Bellman optimality 23 2.3 Rational structure of the reward and an explicit line theorem 26 2.4 Solution methods for Markov decision processes 35 Chapter 3. State-action geometry of partially observable MDPs 44 3.1 The state-action polytope of fully observables systems 45 3.2 State-action geometry of partially observable systems 54 3.3 Number and location of critical points 69 3.4 Reward optimization in state-action space (ROSA) 83 Chapter 4. Geometry and convergence of natural policy gradient methods 94 4.1 Natural gradients 96 4.2 Natural policy gradient methods 101 4.3 Convergence of natural policy gradient flows 107 4.4 Locally quadratic convergence for regularized problems 128 4.5 Discussion and outlook 131 Part II. Neural network-based PDE solvers 133 Chapter 5. Theoretical analysis of the boundary penalty method for neural network-based PDE solvers 134 5.1 Presentation and discussion of the main results 137 5.2 Preliminaries regarding Sobolev spaces and neural networks 146 5.3 Proofs regarding uniform convergence for the deep Ritz method 150 5.4 Proofs of error estimates for the deep Ritz method 156 5.5 Proofs of implications of exact boundary values in residual minimization 167 Chapter 6. Energy natural gradients for neural network-based PDE solvers 174 6.1 Energy natural gradients 176 6.2 Experiments 183 6.3 Conclusion and outlook 192 Bibliography 193