معرفی
Fergus Imrie is a Florence Nightingale Bicentenary Fellow at the University of Oxford's Department of Statistics, affiliated with the Oxford Protein Informatics Group (OPIG) and the Computational Statistics and Machine Learning group. His research focuses on advancing machine learning methodologies for healthcare applications, drug discovery, and predictive modeling with an emphasis on explainability and causal inference. He holds a DPhil from Oxford, supervised by Charlotte Deane, and completed postdoctoral work at UCLA under Mihaela van der Schaar. His work spans generative models for molecular design, multimodal healthcare AI systems, and data-centric methodologies for robust machine learning.
Education
- Doctor of Philosophy (DPhil), University of Oxford (Department of Statistics)
- Postdoctoral Scholar, University of California, Los Angeles (UCLA)
Research Interests
Imrie's methodological contributions include causal inference frameworks, explainable AI techniques, and generative models for structure-based drug design. He develops tools like AutoPrognosis (automated clinical modeling) and MolSnapper (diffusion-based molecular design). Application areas include early disease detection, risk stratification in oncology, and efficient learning from limited/noisy data in drug discovery.
Key Contributions
- Developed ChemIQ benchmark for evaluating LLM chemical intelligence
- Pioneered DC-Check framework for data-centric AI
- Advanced ensemble methods for multimodal healthcare prediction
Grants & Collaborations
Supported by an iCASE award with Exscientia for his doctoral work. Collaborates with leading institutions in computational biomedicine. Advises students on rotation projects in Oxford's DPhil program.
Labs & Teams
Participates in cross-disciplinary teams at OPIG and the van der Schaar Lab, focusing on translational AI solutions for healthcare challenges.
