We invite applications for a fully funded four-year EN & MN Lindsay Endowed PhD studentship in developing in silico approaches for compound optimisation in drug discovery at the University of Dundee, United Kingdom. The student will be based in the School of Life Sciences, one of the UK’s flagship institutions for biological research.
The development of a safe and efficacious drug normally progresses through the optimisation of an initial active compound. During this process the compound undergoes structural changes to improve the desired phenotypic response but also to achieve the required profile in terms of physicochemical (absorption, distribution, metabolism, and excretion) (ADME) and pharmacokinetic properties to be successfully and efficaciously administered to humans. This multi-parameter optimisation (MPO) is very challenging as properties often present conflicting requirements (i.e. changes that are good for one property are detrimental for another). For decades the focus of this optimisation was to obtain compounds as target specific as possible under the assumption that any off-target activity was potentially responsible to undesired side effects. In the last few years, the progress in systems biology has provided a better understanding into the complexity of disease leading to a paradigm shift in drug discovery: compounds acting on multiple targets (polypharmacology) can provide a superior therapeutic effect. A polypharmacology based therapy, where a single drug molecule is able to simultaneously and specifically interact with multiple targets is anticipated to be superior especially in the treatment of complex diseases, such as tumours, nervous system diseases, inflammatory diseases and infectious diseases. In addition, for infectious diseases, it can provide a valuable strategy against the development of drug resistance in anti-infective drugs. However, it faces considerable challenges to how multi-target drugs can be rationally designed. The optimisation of the activity against multiple targets adds an additional level of complexity to the already challenging MPO process typical of drug discovery. Recent advances in machine learning and artificial intelligence (AI) are offering new potential solutions to the MPO challenge by exploring a level of complexity at a scale beyond human capabilities. Machine Learning and Deep Learning methods have been successfully used to predict target activity and molecular properties.1 AI driven generative models have been developed to identify compounds with a desired target activity and properties profile.2 Despite encouraging preliminary results these approaches require further development and validation to have a wider applicability to drug discovery projects. In this project you will work on MPO to design compounds which have drug-like physicochemical properties and the potential to inhibit several different targets using machine learning and AI methods.
1.Walters, W. P., & Barzilay, R. (2021). Applications of Deep Learning in Molecule Generation and Molecular Property Prediction. Accounts of Chemical Research, 54(2), 263–270. https://doi.org/10.1021/acs.accounts.0c00699
2. Blaschke, T., Arús-Pous, J., Chen, H., Margreitter, C., Tyrchan, C., Engkvist, O., … Patronov, A. (2020). REINVENT 2.0: An AI Tool for de Novo Drug Design. Journal of Chemical Information and Modeling, 60(12), 5918–5922. https://doi.org/10.1021/acs.jcim.0c00915
Applicants are expected to hold (or be about to achieve) at least a 2:1 Honours degree in a relevant subject or demonstrably equivalent experience. Previous experience in Computational Chemistry, Machine Learning and programming in Python would be favourable.