PhD on AI-driven multimodal data fusion models in rare diseases KU Leuven in Belgium

KU Leuven’s Department of Human Genetics is offering a PhD position on AI-driven multimodal data fusion models in rare diseases, based in Leuven, Belgium, within the group of Prof. Alejandro Sifrim. The project sits at the intersection of computational biology, artificial intelligence, statistical modelling, spatial omics, precision medicine, and biomedical data integration.

The PhD fellow will develop and benchmark zero-shot multimodal fusion models for rare disease prediction, using melanoma as a use case. The research combines single-cell and spatial multi-omics with histopathology and clinical information to learn robust cross-modal representations for diagnostic prediction. The project aims to advance generative and explainable AI for multimodal data fusion and extend multimodal frameworks with additional omics layers across heterogeneous datasets.

The host environment is highly interdisciplinary and provides access to advanced computational infrastructure, including GPU- and CPU-based high-performance compute servers and large-scale storage. Through LISCO (Leuven Institute for Single-cell Omics), the candidate will also have access to state-of-the-art single-cell and spatial multi-omics technologies, imaging platforms, and bioinformatics resources.

This is a 48-month, full-time doctoral position. The first 36 months are funded by the MSCA Doctoral Network SPACE-MEL, and the appointment is extended to a full four-year doctoral trajectory at KU Leuven. The intended start date is begin 2027. Funding includes a competitive salary with living and mobility allowances, and family allowance where applicable, plus additional KU Leuven benefits such as holiday pay, hospitalization insurance, commuting reimbursement, and access to sports and childcare facilities.

Eligibility requires a master’s degree in Bioinformatics, Computer Science, Artificial Intelligence, Bioscience Engineering, or a related field by the start date, and applicants must be doctoral candidates eligible for the KU Leuven Doctoral School of Biomedical Sciences. The mobility rule applies: candidates must not have lived or carried out their main activity in Belgium for more than 12 months in the 36 months before recruitment. Strong Python programming and deep-learning experience are expected, with PyTorch preferred. A solid machine-learning background is essential, and experience with representation learning, generative models, foundation models, or multimodal integration is a strong advantage. Familiarity with single-cell or spatial transcriptomics, digital pathology, biomedical imaging, Linux, HPC/GPU systems, git, and reproducible workflows is beneficial. Strong English skills are required, and an English test may be requested.

The project includes secondments of 4 months at the University of Manchester and 3 months at Spotlight Pathology Ltd. in the United Kingdom, so willingness to travel and spend extended periods abroad is essential. Applicants should submit their materials through the KU Leuven online job portal only.