Development and application of computational tools for the design of ligand bindingin proteins

点击翻译按钮获取中文摘要

Computational protein design has come a long way in recent years having been used successfully to design new folds, binding interactions and basic catalysts onto existing protein scaffolds. However, the path has been rocky since major achievements had to be questioned and retracted. This has led to a recalibration in the field and it is necessary to refocus on certain aspects of design that we believed to have been solved. We want to concentrate our efforts on the aspect of small molecule ligand interactions and systematically investigate how the design of a binding pocket and its specificity can be improved. The project also has implications for the establishment of catalytic function, since transition state binding is extremely important for the design of specific enzymes. Therefore we want to build our own computational design pipeline and use it for the incremental design in model systems that allow robust experimental validation. For the pipeline we are not only developing tools ourselves, but are also taking advantage of existing well-tested computational methods and algorithms. It will be a modular setup with defined interfaces that allows us to exchange and test different approaches for the various steps in the design process and that can later be extended to other design problems. For experimental validation we use robust biophysical and structural characterization techniques. Further we want to employ experimental library design combined with directed evolution to survey whether the computational design identified is the best solution possible. This systematic analysis will be used as a feedback loop to enhance the computational approach. By this method we hope to implement a reliable design strategy and to contribute significantly to the understanding of protein-ligand interactions.