Improving Rational Drug Design for Membrane Targets with AI and Advanced Simulation
LEAD SUPERVISOR: Professor Philip Biggin, Department of Biochemistry
Co-supervisor: Prof Syma Khalid, Department of Biochemistry
Commercial partner: Vertex Pharmaceuticals, Abingdon
We aim to develop new state-of-the-art tools that will accelerate the discovery of medicines for very difficult and under-explored membrane protein drug targets. The proposal builds upon a very successful previous iCASE studentship between the two partners, but will extend it into new areas of discovery. The timing of this project will also allow an optimal overlap of knowledge transfer between successive students.
In the previous work (currently being written up), we established methodologies for state-of-the-art binding affinity calculations for lipid-exposed binding sites. The work, which breaks new ground, is currently being evaluated within Vertex in terms of how it can be integrated into existing workflows. In this studentship, we wish to explore the development of these tools to compounds that typically act at the membrane-protein-water interface - possibly the most complex solvent environment that a drug can act at, and to date is an area that remains under-explored in terms of drug-design approaches.
Furthermore, this area is particularly timely, as there has been a recent slew of publications, based upon the cryo-EM revolution, that have shown known drugs interacting at sites that sit in the bilayer-water-protein interface environment. As more and more structural information appears (now also substantially supplemented by the success of AlphaFold), it will become essential to characterise the dynamics and conformational properties of these sites in order to optimize the drug design process.
As an example system for the project that is of interest to both Biggin and Vertex, we will focus on a well-known anti-target, the NaV 1.5 sodium channel which is found predominantly in the heart. NaV channels are of particular interest to Vertex (see the recent press release regarding VX-548 - https://investors.vrtx.com/news-releases/news-release-details/vertex-announces-nav18-inhibitor-advancing-phase-2-clinical). As well as being a key anti-target (similar in some ways to hERG), NaV 1.5 is also receiving increased attention for its differing roles in development (in particular there a key difference between embryonic and mature forms of the protein). There is also a substantial amount of data (structural, pharmacological, electrophysiological) both in the public domain as well as within Vertex.
The key aims of the project are to 1) Characterize and understand the problems associated with making accurate absolute binding free energy (affinity) predictions for these types of binding sites 2) To investigate electric-field effects in these calculations and 3) To develop AI-based approaches to speed up key elements of this process – either the calculations per se or predictions without expensive calculations at all.
The results, although focussed on NaV channels will be broadly applicable to many other targets of interest (CFTR, transporters, etc). Novel medicines are both an MRC prioritised area of research (new technologies for advanced therapies) and in the remit of the scientific vision of Vertex. The timing of this synergistic project is particularly appropriate – recent cryo-EM structures of NaV channels, improvements in AI approaches (which we are already incorporating) and the positioning of both partners at the forefront of these areas means that this presents a fantastic opportunity to make significant progress in a difficult field.
Apply using course: DPhil in Biochemistry