Up to three 4-year, fully funded ‘Joint A*STAR – EPSRC DRIVE-Health Studentships’ are available to support PhDs commencing October 2026, covering tuition fees, stipend, and bench fee.
Students recruited to these studentships must spend a minimum of 18 months and a maximum of 24 months at A*STAR Research Institute in Singapore with the named A*STAR supervisor(s) as part of the research and training programme. This is called the “attachment” period, and it will start in their second academic year.
Applications are accepted from citizens of the UK, the EU, the USA, Canada, Latin America, and Australia.
Please read the specific Key Dates and How to Apply sections on this collaboration with A*STAR. Apart from the regular DRIVE-Health entry requirements and application process, A*STAR applicants will have a 2-tier interview process: by a KCL academic panel and a panel from Singapore.
Deep learning tools offer the opportunity to make predictions about complex biological outcomes that cannot be easily extrapolated from experimental manipulations. The goal of this project is to build an AI model that can make predictions about the molecular regulators of muscle stem cells using information about cell behaviour. Muscle weakness in ageing and disease is a costly and debilitating condition that affects quality of life and independent living for many people. Identifying interventions that can enhance muscle function by improving muscle stem cell (muSC) would be greatly beneficial across a wide range of conditions. Cell behaviour offers a rich, spatiotemporal readout of tissue biology that can potentially offer more informative data for predicting outcomes in response to a perturbation and so will form the basis for designing a novel model to predict regulators of muSC function. Models will be developed from time-lapsed videos of muscle stem cells responding to injury, and from spatial maps of gene expression generated from a zebrafish model of regeneration. Deep learning models using computer vision approaches will be developed using time-lapsed movies as input and trained using spatial gene expression as labels. The resultant deep learning model will be used to predict spatial gene expression given a time-lapsed video of cell behaviour. This project builds on the expertise of the Knight and Lee groups in cell and tissue biology, spatial transcriptomics and deep learning this project offers a unique opportunity to explore the utility of AI models for investigating a challenging area of health.