PhD Position in Large Language Models for Multimodal Spatial Omics in Melanoma

This doctoral position focuses on developing and implementing Large Language Model (LLM) architectures for the natural-language exploration of multimodal spatial embeddings in melanoma research. Hosted by Intelligent Biodata SL in Spain and co-supervised by the University of Navarra, the successful candidate will fine-tune pretrained models for spatial multi-omics and build accessible interfaces for clinical users. The project includes international secondments at KU Leuven in Belgium and Spotlight Pathology Ltd in the United Kingdom. The candidate is expected to conduct collaborative research, participate in network training events, and contribute to precision medicine software tools.

• Benchmark and adapt existing large language model architectures for spatial multi-omics applications
• Fine-tune pretrained language models using public datasets and consortium-generated melanoma data
• Develop accessible graphical interfaces that enable clinical and research users to explore integrated datasets without advanced computational expertise
• Participate in a six-month research secondment at KU Leuven focusing on multimodal data fusion and representation learning
• Complete a three-month secondment at Spotlight Pathology Ltd to focus on user-friendly software design for clinical deployment
• Collaborate with international academic and industrial partners within the SPACE-MEL network
• Present research findings at international workshops, courses, and conferences