Harnessing AI to Define Extracellular Vesicle-Mediated Immune Regulation

This PhD project at Queen’s University Belfast, within the School of Medicine, Dentistry & Biomedical Sciences, explores the molecular mechanisms by which extracellular vesicles (EVs) derived from mesenchymal stem cells (MSCs) regulate immune cell function. EVs are nanoscale particles released by most cell types, including immune and stem cells, and are known to carry a diverse array of molecular cargo such as proteins, lipids, and nucleic acids. These vesicles play a pivotal role in immune communication, modulating activation, differentiation, and effector functions of recipient cells. Despite their recognized importance, the precise molecular mechanisms by which EVs influence immune responses remain incompletely understood. The project leverages advanced artificial intelligence (AI) and machine learning (ML) techniques to integrate multi-omics datasets, including high-resolution mass spectrometry and RNA sequencing, to profile the cargo of MSC-derived EVs. EVs will be isolated from cultured MSCs under both resting and activated conditions using size-exclusion chromatography, with their size and purity assessed by nanoparticle tracking analysis. Purified EVs will be used to treat various immune cell populations, including dendritic cells, B cells, and T lymphocytes. The responses of these recipient cells will be quantified using single-cell transcriptomics and flow cytometry. AI/ML models, such as graph neural networks and multimodal autoencoders, will be developed to integrate EV cargo data with immune cell perturbation responses, aiming to identify molecular features predictive of functional outcomes. Interpretable ML methods (e.g., SHAP, network inference analyses) will be employed to highlight key EV molecules or cargo combinations driving immune modulation. Experimental validation will involve perturbing MSC cargo through gene silencing or metabolic modulation to establish causal mechanisms. This integrative approach is designed to move beyond descriptive correlations, providing mechanistic insights into how specific molecular signatures within MSC-derived EVs regulate immune cell behaviour. The project is funded by the BBSRC for four years and is open to both Home and International applicants. However, only a small number of NILAB awards are available for international students, and these will be allocated competitively across all NILAB projects based on the strength of applications. Applicants should have a strong background in mathematics, computer science, bioinformatics, or a related quantitative discipline. Experience with Python, machine learning, or probabilistic modelling is advantageous but not essential. The application deadline is February 16, 2026. Interested candidates should apply online via the FindAPhD project page and prepare their application materials accordingly. For further details, visit the project website.