I obtained my MSci in chemistry from Imperial College London from 2013 - 2017, culminating in a DFT-based research project. I then went on to do an MSc in scientific computing, including projects in machine learning, high performance computing and applied mathematics. In 2018, I started on the BBSRC Interdisciplinary Biosciences DTP, with research in the Oxford Protein Informatics Group (OPIG) beginning in January 2019. Research is focussed on using machine learning to probe small molecules and their interactions with proteins. Funding is through the iCASE scheme, partnered with BenevolentAI.
For my four month MSc project, I overhauled the way in which voronoi diagrams are calculated in the Voronoi Image Segmentation algorithm (VOISE). By changing the language from Matlab to C++ (MEX) and implementing the SKIZ Operator Algorithm for dynamic local recalculation of Voronoi regions, I affected a ~240 fold increase in speed. I also modified and implemented a version of the kNN-enhance algorithm for community detection, using the resulting Delaunay triangulations as input graphs to perform clustering on the output of VOISE, with improved results when compared to the previously used k-means algorithm.
Completed a nine month long research project for my MSci in chemistry at Imperial College London, supervised by Professor Nicholas Harisson. The antiferromagnetic coupling in twice oxidised transition metal phthalocyanines (MPcs) was predicted using DFT, showing good agreement with experimentally observed values for those which had been experimentally characterised. This model was then used to predict band structure and intermolecular spin coupling in thin films of as-yet uncharacterised MPcs.