Quantitative approaches for characterising fibrillar protein nanostructures

Polypeptide sequences have an inherent tendency to self-assemble into filamentous nanostructures commonly known as amyloid fibrils. Such self-assembly is used in nature to generate a variety of functional materials ranging from protective coatings in bacteria to catalytic scaffolds in mammals. The aberrant self-assembly of misfolded peptides and proteins is also, however, implicated in a range of disease states including neurodegenerative conditions such as Alzheimer's and Parkinson's diseases. It is increasingly evident that the intrinsic material properties of these structures are crucial for understanding the thermodynamics and kinetics of the pathological deposition of proteins, particularly as the mechanical fragmentation of aggregates enhances the rate of protein deposition by exposing new fibril ends which can promote further growth. We discuss here recent advances in physical techniques that are able to characterise the hierarchical self-assembly of misfolded protein molecnles and define their properties. © 2010 Materials Research Society.

Probabilistic inference for fast learning in control

A novel framework is provided for very fast model-based reinforcement learning in continuous state and action spaces. It requires probabilistic models that explicitly characterize their levels of condence. Within the framework, exible, non-parametric models are used to describe the world based on previously collected experience. It demonstrates learning on the cart-pole problem in a setting where very limited prior knowledge about the task has been provided. Learning progressed rapidly, and a good policy found after only a small number of iterations.