Phd Position In Large Language Models For Multimodal Spatial Omics In Melanoma — Space-Mel Dc9

Organisation/Company intelligent biodata Research Field Computer science » Other Biological sciences » Other Researcher Profile First Stage Researcher (R1) Positions PhD Positions Application Deadline 30 Oct 2026 - 12:00 (Europe/Madrid) Country Spain Type of Contract Temporary Job Status Full-time Is the job funded through the EU Research Framework Programme? Horizon Europe - MSCA Reference Number SPACE-MEL — DC9 — GA 101311007 Is the Job related to staff position within a Research Infrastructure? No

Offer Description

Academic co-supervisor: Prof. Ángel Rubio, University of Navarra, TECNUN School of Engineering, San Sebastián, Spain.

Additional co-supervisors and secondment hosts: Prof. Alejandro Sifrim, KU Leuven, Department of Human Genetics, Belgium (Academia);
Dr. Samantha Perona, Spotlight Pathology Ltd., United Kingdom (Industry).

DC9 will develop and implement a Large Language Model (LLM)-based system for natural-language exploration of multimodal spatial embeddings in melanoma research. The project will use spatial multi-omics, histology-derived features and consortium-generated datasets to support intuitive querying and interpretation of complex melanoma data.

The project pursues two integrated objectives: (i) fine-tune pretrained LLMs for melanoma-relevant spatial multi-omics data;
and (ii) develop an accessible interface that enables clinical and research users to explore integrated datasets without advanced computational expertise.

At Intelligent Biodata, the candidate will benchmark and adapt existing LLM architectures for spatial multi-omics applications, using public datasets and SPACE-MEL data to improve model accuracy and relevance. During a first secondment with Prof. Alejandro Sifrim at KU Leuven, the candidate will receive training in multimodal data fusion and representation learning. A second secondment at Spotlight Pathology Ltd., supervised by Dr. Samantha Perona, will focus on developing a user-friendly graphical interface for clinical and research users.

Expected outcomes include: (1) an LLM for spatial multi-omics;
(2) integrated melanomadatasets;
and (3) an accessible software tool for non-computational users.

Host institution and doctoral training

The doctoral candidate will be employed by Intelligent Biodata SL in Spain under the industrial supervision of Dr. Jon Pey. The academic component of the doctoral training will be conducted in collaboration with the University of Navarra, particularly the TECNUN School of Engineering in San Sebastián, under the academic co-supervision of Prof. Ángel Rubio.

The candidate will benefit from complementary industrial and academic environments, with activities at both Intelligent Biodata and the university.

Intelligent Biodata is a bioinformatics company with expertise in artificial intelligence, machine learning, natural language processing, multimodal data integration, bioinformatics pipelines and software development for biomedical applications. The company develops computational methods and deployable software solutions for the analysis of complex multi-omics, imaging and clinical datasets.

The host offers access to development and production infrastructure for AI model implementation, testing, deployment and monitoring, including GPU resources, cloud computing and secure data-transfer infrastructure with KU Leuven. The candidate will benefit from expertise in model development, biomedical data integration, software engineering and deployment of AI solutions for precision medicine.

Secondments

The project includes research and training visits to the following partner organisations. A willingness to travel and spend time abroad is essential:

• Six months at KU Leuven, Department of Human Genetics, Belgium, supervised by Prof. Alejandro Sifrim.
• Three months at Spotlight Pathology Ltd., United Kingdom, supervised by Dr. Samantha Perona.

The SPACE-MEL consortium

SPACE-MEL is an EU-funded MSCA Doctoral Network advancing spatial single-cell omics technologies for precision medicine. The consortium brings together European universities, technology providers, clinical partners and companies to address experimental and computational challenges limiting the clinical translation of spatial biology.

By integrating multi-omics approaches, enabling the transition from 2D to 3D tissue analysis and developing advanced data interpretation methods, SPACE-MEL aims to improve our understanding of cellular organisation and disease mechanisms.

The network will train 15 doctoral candidates through an interdisciplinary programme spanning spatial imaging technologies, molecular biology, computational biology, pathology and biomarker discovery. Cross-sectoral secondments, hands-on research and transferable-skills training will provide experience across the spatial biology workflow.

What SPACE-MEL offers