Search

High-throughput in vitro screening and crystal structures identify a non-statin inhibitor of HMG-CoA reductase for novel antibacterial drug design. HMG-CoA reductase (HMGR), a rate-limiting enzyme of the mevalonate pathway in Gram-positive pathogenic bacteria, is an attractive target for development of novel antibiotics. In this study, we report the crystal structures of HMGR from Enterococcus faecalis (efHMGR) in the apo and liganded forms, highlighting several unique features of this enzyme. Statins, which inhibit the human enzyme with nanomolar affinity, perform poorly against the bacterial HMGR homologs. We also report a potent competitive inhibitor (Chembridge2 ID 7828315 or compound 315) of the efHMGR enzyme identified by a high-throughput, in-vitro screening. The X-ray crystal structure of efHMGR in complex with 315 was determined to 1.27 angstrom resolution revealing that the inhibitor occupies the mevalonate-binding site and interacts with several key active site residues conserved among bacterial homologs. Importantly, 315 does not inhibit the human HMGR. Our identification of a selective, non-statin inhibitor of bacterial HMG-CoA reductases will be instrumental in lead optimization and development of novel antibacterial drug candidates.

We discuss the O(2N) vector model in three dimensions. While this model flows to the Wilson-Fisher fixed point when fine tuned, working in a double-scaling limit of large N and large charge allows us to study the model away from the critical point and even to follow the RG flow from the UV to the IR. The crucial observation is that the effective potential -- at leading order in N but exact to all orders in perturbation theory -- is the Legendre transform of the grand potential at fixed charge. This allows us to write an effective action and the free energy for generic values of the coupling in a very simple fashion and without evaluating any Feynman diagrams., Comment: 20 pages, 4 figures

We discuss the O(2N) vector model in three dimensions. While this model flows to the Wilson-Fisher fixed point when fine tuned, working in a double-scaling limit of large N and large charge allows us to study the model away from the critical point and even to follow the RG flow from the UV to the IR. The crucial observation is that the effective potential -- at leading order in N but exact to all orders in perturbation theory -- is the Legendre transform of the grand potential at fixed charge. This allows us to write an effective action and the free energy for generic values of the coupling in a very simple fashion and without evaluating any Feynman diagrams., Comment: 20 pages, 4 figures

The design of embedded systems is a well-established research domain for many decades. However, the constantly increasing complexity and requirements of state-of-the-art embedded systems pushes designers to new challenges while maintaining established design methodologies. Embedded system design uses the concept of Discrete Event Simulation (DES) to prototype and test the interaction of individual components.In this dissertation, we provide the Recoding Infrastructure for SystemC (RISC) compiler framework to perform static and hybrid analysis of IEEE SystemC models. On one hand, RISC generates thread communication charts to visualize the communication between individual design components. The visualization respects the underlying discrete event simulation semantics and illustrates the individual synchronization steps. On the other hand, RISC translates a sequential model into a parallel model which effectively utilizes multi- and many-core host simulation platforms for faster simulation. This work extends the conflict analysis capabilities for libraries, dynamic memory allocation, channel instance awareness, and references. Additionally, the traditional thread level parallelism is extended with data level parallelism for even faster simulation.

The design of embedded systems is a well-established research domain for many decades. However, the constantly increasing complexity and requirements of state-of-the-art embedded systems pushes designers to new challenges while maintaining established design methodologies. Embedded system design uses the concept of Discrete Event Simulation (DES) to prototype and test the interaction of individual components.In this dissertation, we provide the Recoding Infrastructure for SystemC (RISC) compiler framework to perform static and hybrid analysis of IEEE SystemC models. On one hand, RISC generates thread communication charts to visualize the communication between individual design components. The visualization respects the underlying discrete event simulation semantics and illustrates the individual synchronization steps. On the other hand, RISC translates a sequential model into a parallel model which effectively utilizes multi- and many-core host simulation platforms for faster simulation. This work extends the conflict analysis capabilities for libraries, dynamic memory allocation, channel instance awareness, and references. Additionally, the traditional thread level parallelism is extended with data level parallelism for even faster simulation.

Covert channels are communication channels used by attackers to transmit information from a compromised system when the access control policy of the system does not allow doing so. Previous work has shown that CPU frequency scaling can be used as a covert channel to transmit information between otherwise isolated processes. Modern systems either try to save power or try to operate near their power limits in order to maximize performance, so they implement mechanisms to vary the frequency based on load. Existing covert channels based on this approach are either easily thwarted by software countermeasures or only work on completely idle systems. In this paper, we show how the automatic frequency scaling provided by Intel Turbo Boost can be used to construct a covert channel that is both hard to prevent without significant performance impact and can tolerate significant background system load. As Intel Turbo Boost selects the maximum CPU frequency based on the number of active cores, our covert channel modulates information onto the maximum CPU frequency by placing load on multiple additional CPU cores. Our prototype of the covert channel achieves a throughput of up to 61 bit/s on an idle system and up to 43 bit/s on a system with 25% utilization.

The Program State Machine (PSM) Model of Computation offers a rich set of modeling elements to describe behavioral and structural hierarchy, concurrency, synchronization, state transitions and timing. With the rising software complexity of today's embedded systems, the use of Real-Time Operating Systems (RTOS) has become state-of-the-art for nearly all System-on-Chip designs. Regrettably, the PSM model itself has insufficient support for the specification of the preemptive dynamic scheduling behavior of an RTOS. In this paper, we propose a model for dynamically dispatching PSM models on a virtual processing element. Our model aims to abstract from the targeted RTOS and the processor core through execution time annotations and a flexible preemptive scheduler model. Mapping a PSM model to a set of scheduled virtual processing elements only requires minor model transformation and enables early exploration of different processing element mappings and scheduling policies. Our virtual processing model for PSMs is realized on top of the SystemC library. We evaluate the proposed virtual processing model using a Canny edge detection filter.

The Program State Machine (PSM) Model of Computation offers a rich set of modeling elements to describe behavioral and structural hierarchy, concurrency, synchronization, state transitions and timing. With the rising software complexity of today's embedded systems, the use of Real-Time Operating Systems (RTOS) has become state-of-the-art for nearly all System-on-Chip designs. Regrettably, the PSM model itself has insufficient support for the specification of the preemptive dynamic scheduling behavior of an RTOS. In this paper, we propose a model for dynamically dispatching PSM models on a virtual processing element. Our model aims to abstract from the targeted RTOS and the processor core through execution time annotations and a flexible preemptive scheduler model. Mapping a PSM model to a set of scheduled virtual processing elements only requires minor model transformation and enables early exploration of different processing element mappings and scheduling policies. Our virtual processing model for PSMs is realized on top of the SystemC library. We evaluate the proposed virtual processing model using a Canny edge detection filter.

Artificial Intelligence as a field concerns itself with the study and design of intelligent agents. Such agents or systems perceive their environment, reason over their accumulated perceptions and through this reasoning, derive a course of action which achieves their goals or maximizes their performance measure. In many cases this planning process boils down to a principled evaluation of potential sequences of actions by exploring the resulting anticipated situations of the agent and its environment, that is to some form of combinatorial search in an implicit graph representing the interdependencies of agent actions and potential situations. This thesis focuses on the search algorithms that implement this reasoning process for intel- ligent agents. The first part covers a novel subclass of such reasoning problems where a goal driven agent must find a sequence of actions to achieve its goal such that some value function of the sequence is closest to a reference value. Such problems occur for example in recom- mender systems and self-diagnosing agents. The contribution here is an algorithm and a family of heuristic functions based on memoization that improves on the current state-of-the-art by several orders of magnitude. The second part focuses on cost- and step-optimal planning for goal-driven agents in which the problem is to derive the least-cost action sequence which achieves the agent’s goals. Here, dynamic programming can be employed to avoid redundant evaluations. The contributions of this thesis address two challenges associated with the use of dynamic programming algo- rithms. The first challenge is the memory consumption of these techniques. At the heart of such search algorithms lies a memoization component that keeps track of all generated situations and the best known way to attain them which is used to derive potential new situations. In practice the rapid growth of this component is the limiting factor for the applicability of this class of algorithms, Intelligente Software-Agenten zeichnen sich durch die Fähigkeit aus durch Schlussfolgerungsprozesse und Umgebungswissen eigenständig Aktionen auszuwählen und durchzuführen, die geeignet sind ihre vorgegebenen Ziele zu erreichen. In vielen Fällen sind diese Schlussprozesse im Kern als heuristische Suche über einem Umgebungs- und Aktionsmodell implementiert, wobei die Skalierbarkeit dieser Suche sowohl die mögliche Umgebungskomplexität als auch die Qualität des Agentenverhaltens direkt beeinflusst. Diese Dissertation greift dieses Skalierungsproblem auf und beschreibt im ersten Teil ein neuartiges Suchverfahren für eine Klasse von komplexen Suchproblemen die im Kontext selbstdiagnostizierender Agenten auftreten und im zweiten Teil eine neuartige Datenstruktur und Parallelisierungstechnik für auf Dynamischer Programmierung basierender Suchverfahren, die ihren Einsatz in komplexeren Umgebungen ermöglicht.

Artificial Intelligence as a field concerns itself with the study and design of intelligent agents. Such agents or systems perceive their environment, reason over their accumulated perceptions and through this reasoning, derive a course of action which achieves their goals or maximizes their performance measure. In many cases this planning process boils down to a principled evaluation of potential sequences of actions by exploring the resulting anticipated situations of the agent and its environment, that is to some form of combinatorial search in an implicit graph representing the interdependencies of agent actions and potential situations. This thesis focuses on the search algorithms that implement this reasoning process for intel- ligent agents. The first part covers a novel subclass of such reasoning problems where a goal driven agent must find a sequence of actions to achieve its goal such that some value function of the sequence is closest to a reference value. Such problems occur for example in recom- mender systems and self-diagnosing agents. The contribution here is an algorithm and a family of heuristic functions based on memoization that improves on the current state-of-the-art by several orders of magnitude. The second part focuses on cost- and step-optimal planning for goal-driven agents in which the problem is to derive the least-cost action sequence which achieves the agent’s goals. Here, dynamic programming can be employed to avoid redundant evaluations. The contributions of this thesis address two challenges associated with the use of dynamic programming algo- rithms. The first challenge is the memory consumption of these techniques. At the heart of such search algorithms lies a memoization component that keeps track of all generated situations and the best known way to attain them which is used to derive potential new situations. In practice the rapid growth of this component is the limiting factor for the applicability of this class of algorithms, Intelligente Software-Agenten zeichnen sich durch die Fähigkeit aus durch Schlussfolgerungsprozesse und Umgebungswissen eigenständig Aktionen auszuwählen und durchzuführen, die geeignet sind ihre vorgegebenen Ziele zu erreichen. In vielen Fällen sind diese Schlussprozesse im Kern als heuristische Suche über einem Umgebungs- und Aktionsmodell implementiert, wobei die Skalierbarkeit dieser Suche sowohl die mögliche Umgebungskomplexität als auch die Qualität des Agentenverhaltens direkt beeinflusst. Diese Dissertation greift dieses Skalierungsproblem auf und beschreibt im ersten Teil ein neuartiges Suchverfahren für eine Klasse von komplexen Suchproblemen die im Kontext selbstdiagnostizierender Agenten auftreten und im zweiten Teil eine neuartige Datenstruktur und Parallelisierungstechnik für auf Dynamischer Programmierung basierender Suchverfahren, die ihren Einsatz in komplexeren Umgebungen ermöglicht.

In development, tissue regeneration or certain diseases, angiogenic growth leads to the expansion of blood vessels and the lymphatic vasculature. This involves endothelial cell proliferation as well as angiogenic sprouting, in which a subset of cells, termed tip cells, acquires motile, invasive behaviour and extends filopodial protrusions. Although it is already appreciated that angiogenesis is triggered by tissue-derived signals, such as vascular endothelial growth factor (VEGF) family growth factors, the resulting signalling processes in endothelial cells are only partly understood. Here we show with genetic experiments in mouse and zebrafish that ephrin-B2, a transmembrane ligand for Eph receptor tyrosine kinases, promotes sprouting behaviour and motility in the angiogenic endothelium. We link this pro-angiogenic function to a crucial role of ephrin-B2 in the VEGF signalling pathway, which we have studied in detail for VEGFR3, the receptor for VEGF-C. In the absence of ephrin-B2, the internalization of VEGFR3 in cultured cells and mutant mice is defective, which compromises downstream signal transduction by the small GTPase Rac1, Akt and the mitogen-activated protein kinase Erk. Our results show that full VEGFR3 signalling is coupled to receptor internalization. Ephrin-B2 is a key regulator of this process and thereby controls angiogenic and lymphangiogenic growth.