Investigations of Chemical Space using Machine Learning

In this project modern machine learning (e.g. deep learning, transfer learning and reinforcement learning) aspects of artificial intelligence will be used to model and investigate areas of chemical space in order to develop new molecular representations that facilitate our ability to direct attention to the regions of the chemical space that are relevant to the discovery of new active compounds, for use for example as antibiotics. Similar techniques will also be applied to use ML in computationally assisted materials and molecular imaging.