Machine Learning

Machine Learning

Philip Fowler

Part of the time I lead a small group of researchers and together we are developing methods that can predict whether mutations confer resistance to antibiotics or not. I’m also involved in GPAS and worked for several years on tuberculosis, notably the CRyPTIC and BashTheBug projects. This broad range reflects my background as an interdisciplinary scientist — I originally trained as a physicist, gained a PhD in chemistry, worked in Biochemistry and am now based in the John Radcliffe hospital.  

Alice Brankin

I am a PhD student, I work on several projects relating to antibiotic resistance in tuberculosis. I use computer programming to analyse genetic and drug resistance data from the Comprehensive Resistance Prediction for Tuberculosis: an International Consortium (CRyPTIC) project and use machine learning and computational chemistry to predict resistance associated with different genetic mutations. To become a PhD student in this area, I did undergraduate and Masters degrees in biochemistry, had an interest in programming and was motivated to continue doing research. For my work with CRyPTIC I am responsible for conducting data analysis, making visualisations and presenting our research in the form of manuscripts and talks.

Charlotte Lynch

I work on predicting antimicrobial resistance in Tuberculosis. To do this I combine machine learning computer models with genetics-based data (via the CRyPTIC dataset) and experimental data on the structure of protein molecules. I joined the project via a rather unusual route – I was originally in Materials Science where I was using quantum mechanics to predict the properties of materials. Then I moved to Biochemistry where I used classical mechanics to predict the behaviour of proteins. Now I’m a postdoc in the Fowler Lab within the Modernising Medical Microbiology group where I’m building a set of computational tools to predict antimicrobial resistance.