280 Projects, page 1 of 56
We propose to computationally design novel ligand binding and catalytically active proteins by harnessing the high thermodynamic stability of de novo helical proteins. Tremendous progress has been made in protein design. However, the ability to robustly introduce function into genetically encodable de novo proteins is an unsolved problem. We will follow a highly interdisciplinary computational-experimental approach to address this challenge and aim to: -Characterize to which extent we can harness the stability of parametrically designed helical bundles to introduce deviations from ideal geometry. Ensembles of idealized de novo helix bundle backbones will be generated using our established parametric design code and designed with constraints accounting for an envisioned functional site. This will be followed by detailed computational, biophysical, crystallographic and site-saturation mutagenesis analysis to isolate critical design features. -Develop a new computational design strategy, which expands on the Crick coiled-coil parametrization and allows to rationally build non-ideal helical protein backbones at specified regions in the desired structure. This will enable us to model backbones around binding/active sites. We will design sites to bind glyphosate, for which remediation is highly needed. By using non-ideal geometries and not relying on classic heptad repeating units, we will be able to access a much larger sequence to structure space than is usually available to nature, enabling us to build more specific and more stable binding/catalytically active proteins. -Investigate new strategies to design the first cascade reactions into de novo designs. This research will allow functionalization of de novo designed proteins with high thermostability, extraordinary resistance to harsh chemical environments and high tolerance for organic solvents and has the potential to revolutionize how proteins for biotechnological and biomedical applications are generated.
Energy storage is undeniably amongst the greatest societal challenges. Batteries will be key enablers but require major progress. Battery materials that promise a step-change in energy density compared with current Li-ion batteries rely on fundamentally different reactions to store charge, e.g. Si alloying or O2 reduction instead of intercalation. They have in common high volume changes on cycling and poor conductivity. For the active component of a battery electrode to function it must be simultaneously in contact with ionic and electronic pathways to electrolyte and current collector. State-of-the-art conducting additives and binders in the composite electrodes cannot ensure ideal contact for such materials and fail to exploit their full potential. In this project I directly target these fundamental challenges of high-energy batteries by replacing now used conducting additives and binders with flexible organic mixed ion and electron conductors that follow volume changes to ensure at any stage intimate contact with ions and electrons. This requires progress with the fundamental science of such conductors, for which to achieve we develop and combine synthetic, electroanalytic and spectroscopic methods, aided by theory. Mixed conducting polymer gels, designed for the particular storage material, shall be elaborated for two ultra-high capacity electrodes, the O2 cathode and the Si anode. The significant advantage, next to intimate contact, is that the packing density of active material can be maximized. This boosts energy stored by total electrode mass and volume by rigorously cutting the amount of non-active materials compared with current approaches. The expected overriding scientific impact includes improved understanding of mixed conductors concerning synthesis, structure, conductivity and their behaviour in the complex battery environment. This opens up new perspectives for the realm of high-capacity battery materials that demand such a breakthrough to succeed
More than 15 years ago, several seminal publications showed that cryptographic keys can be revealed by analysing the power consumption or by inducing faults to devices like smart cards. The publication of these so-called physical attacks sparked off research on all kinds of attack techniques and countermeasures to secure implementations of cryptographic schemes. However, a system can still be attacked easily if only the execution of cryptographic schemes is secured. An attacker can for example induce a fault to bypass an authentication or to jump to a privileged function directly. The system might also leak the key before the execution of a cryptographic scheme starts. Today, there is almost no research on securing systems and software execution against physical attacks. Products like smart cards rely on proprietary best-practice countermeasures. Also countless devices of the Internet of Things are exposed to physical attacks and lack protection. Our goal is to close this fundamental gap in system security and to establish the scientific foundation for executing software securely and efficiently in the presence of physical attacks. We aim to address research questions that range from the modelling of the attacks at the hardware level up to system-level questions like how changing properties of programming languages can support achieving protection against physical attacks. This project brings together research on physical attacks, cryptography, system architectures, fault tolerant design as well as formal methods. Combining the fields, we pursue novel approaches to securing the control flow, CPU computations and memories. We in particular aim to find efficient methods in hardware and software that allow building systems where critical parts of the overall software can be secured against physical attacks without affecting or trusting the rest of the system. Our research also includes automated generation and verification techniques for the secured software.
Since more than 50 years, computer vision has been a very active research field but it is still far away from the abilities of the human visual system. This stunning performance of the human visual system can be mainly contributed to a highly efficient three-layer architecture: A low-level layer that sparsifies the visual information by detecting important image features such as image gradients, a mid-level layer that implements disocclusion and boundary completion processes and finally a high-level layer that is concerned with the recognition of objects. Variational methods are certainly one of the most successful methods for low-level vision. However, it is very unlikely that these methods can be further improved without the integration of high-level prior models. Therefore, we propose a unified mathematical framework that allows for a natural integration of high-level priors into low-level variational models. In particular, we propose to represent images in a higher-dimensional space which is inspired by the architecture for the visual cortex. This space performs a decomposition of the image gradients into magnitude and direction and hence performs a lifting of the 2D image to a 3D space. This has several advantages: Firstly, the higher-dimensional embedding allows to implement mid-level tasks such as boundary completion and disocclusion processes in a very natural way. Secondly, the lifted space allows for an explicit access to the orientation and the magnitude of image gradients. In turn, distributions of gradient orientations – known to be highly effective for object detection – can be utilized as high-level priors. This inverts the bottom-up nature of object detectors and hence adds an efficient top-down process to low-level variational models. The developed mathematical approaches will go significantly beyond traditional variational models for computer vision and hence will define a new state-of-the-art in the field.