Towards a Cognitive Assistant for Computational Chemistry: Investigating automatable methods to analyse the output of simulations (iCase joint with IBM Research UK)


Description: With the emergence of high-throughput virtual screening as a tool for materials discovery, larger and more complex data sets are being produced as the output from simulation. Indeed, in the foreseeable future, the human task of identifying and classifying distinguishing features from the data will become the bottleneck to progress. This project will investigate automatable methods to analyze the output of simulations, extract relevant features, and use these features to both build on the fly QSPR type predictors, and to suggest new simulations which could bring additional value. A wide range of simulation classes will be analysed, from coarse grained models of bulk materials, to atomistic methods for small molecules, and extending into QM methods for investigating electronic structure. 

Environment: This studentship will be based for at least 6 months at IBM Research UK within the Hartree Center in Daresbury, which was establishged to transform the competitiveness of UK industry by accelerating the adoption of High Performance Computing, Big Data and Cognitive technologies. Other Hartree focus areas include high accuracy formulation in consumer goods, manufacturing challenges and life sciences projects such as precision agriculture, anti-microbial surfaces and genomics. At the University, the studentship will be based in the Materials Innovation Factory (MIF), a new £68 M research facility, supervised by Prof. A. I. Cooper FRS, the MIF Academic Director. The studentship is funded by EPSRC but will also form a part of the Leverhulme Research Centre for Functional Materials Design, a new £10 M, 10-year activity funded by the Leverhulme Trust. 

Qualifications: A 2:1 or higher degree or equivalent in Chemistry with a strong interest in data-science, or alternatively a strong interest in physical science but with Mathematics or Computer Science background. The candidate will be expected to have strong programming abilities (Python preferred), and an interest in the application of machine-learning techniques to complex chemical problems. 

Informal enquiries should be addressed to Professor Cooper 

To apply for this opportunity, click here.