Designing Plant Gene Expression with Explainable AI (LI_E27DTP)

Plants survive drought, heat and disease by rapidly changing which genes are switched on, where, and when. But we still cannot reliably design the regulatory DNA that controls these expression patterns. This PhD asks an ambitious question: can explainable AI discover plant DNA “design rules” that are strong enough to survive experimental testing?

You will join Earlham Institute’s AI for Biology group led by Professor Ke Li. As one of the most active AI groups in the UK, you will work in a highly supportive and collaborative group, equipped with unique technical platforms for building in-house large-scale foundation models. You will receive scientific training in foundational AI methods and push the boundary of AI for biology. Meanwhile, you will also develop skills in experimental design, reproducible software and FAIR data management.

You will also obtain complementary experience in plant genomics and engineering biology, supported by Prof. Anthony Hall’s plant genomics expertise and Dr Carolina Grandellis’s Earlham Biofoundry team. You will collaborate with experimentalists on the design of compact libraries of cis-regulatory modules, assemble them into a standard reporter system and test them in potato protoplast assays. Successful and failed designs will both be valuable: they will reveal which AI-derived rules are biologically meaningful, where models fail, and how future design cycles should be improved.

The project will be conducted at the Earlham Institute, a world-class BBSRC research centre of excellence in genomics, engineering and computational biology, and AI. You will join a highly collaborative and supportive research community, interact with AI researchers, plant scientists, bioinformaticians, automation specialists, and have access to training and career development opportunities across the Norwich Research Park through the Norwich Research Park Biosciences Doctoral Training Partnership.