The use of machine learning models to identify new non-nucleoside reverse transcriptase inhibitors targeting HIV-1

Recent developments in machine learning have stimulated interest in the application of this powerful technique to new areas of research, including drug discovery. The increased availability of diverse electronic data sources means that drug discovery is undergoing a data-driven shift in which more traditional computational approaches can be augmented using machine learning protocols for enhanced in-silico screening [1].

This project will look to use various machine learning approaches, including deep learning neural models and Bayesian models to develop predictive models for identifying potential leads to target HIV-1 reverse transcriptase (RT) in an allosteric binding mode [2]. The leads identified will then be verified using traditional molecular modelling, synthesised, and evaluated using a range of biological screening methods.

The multidisciplinary nature of this project represents a great opportunity for the student and would be particularly suited to a Chemistry/Medicinal Chemistry or Pharmacy graduate and allow them to obtain valuable training in organic chemistry, statistical analysis, medicinal chemistry, and computational modelling. Previous experience with computer programming would be advantageous but is not essential.

The candidate should have (or expect to receive) a degree at 2:1 or above.