PhD Studentship: Data-Driven Realisation of Molecular Editing for Drug Discovery

Funding

EPSRC CASE Competition Studentship in partnership with Exscentia Ltd, offering the award of fees, together with a tax-free maintenance grant of £19,237 and an additional Top-Up of £3,300 per year for 3.5 years.  Training and support will also be provided.

Lead Supervisor’s full name & email address

Professor Adam Nelson – a.s.nelson@leeds.ac.uk

Co-supervisor name(s)

Professor Steve Marsden – s.p.marsden@leeds.ac.uk

Project summary 

Drug discovery pipelines are driven by iterative cycles in which molecules are designed, synthesised, purified and tested.  A remarkably limited toolkit of reactions dominates discovery, which has contributed to the historic uneven exploration of chemical space, and has tended to focus attention on molecules with sub-optimal properties.  Many reactions that would be potentially valuable for drug discovery have recently emerged that could complement this established reaction toolkit.

The main hindrance in harnessing a broader reaction toolkit, such as molecular editing reactions, stems from insufficient knowledge of applicability across a range of substrates, and, thus, a low confidence in using these methods in a resource-pressured real-world context. Synthetic challenges can arise because bioactive molecules are typically more highly functionalised and relatively polar, and such substrates systematically perform less well in reactions that have been optimised using model (simple, commercially available) substrates.  How, then, can the reaction toolkit be broadened to enable molecular editing reactions to be harnessed within drug discovery programmes?

We propose to develop a data-driven approach to enable prediction of the success of molecular editing reactions.  The specific reactions to be investigated will be chosen on the basis of potential strategic value for drug discovery e.g. skeletal editing reactions that enable a core ring system to be precisely and directly modified (vide infra).   Initially, we will establish high-throughput methods to assemble training data by determination of reaction outcomes as a function of the substrates and conditions used.  We will develop machine learning strategies to enable prediction of the outcome of reactions outside the training set.  Finally, we will validate the approach by comparing the predicted and experimentally-determined outcomes of a range of molecular editing reactions involving substrates outside the training set. Overall, the resulting tools will enable uptake of molecular editing reactions within drug discovery.

The project is collaborative with Exscientia, an AI-enabled drug discovery company.

References

None

Please state your entry requirements plus any necessary or desired background

First or Upper Second Class UK Bachelor (Honours) or equivalent

Subject Area

Synthetic Chemistry, AI and Machine Learning

Advert information