Building an open knowledge graph for targeted radionuclide therapy: integrating public data

Targeted Radionuclide Therapy (TRT) is a promising approach in nuclear medicine that delivers therapeutic radiation directly to tumor cells through radiolabeled molecules. At SCK CEN, ongoing research aims to develop new TRT strategies for glioblastoma, ovarian, and colorectal cancers. To support this effort from a computational perspective, this thesis will explore how publicly available biomedical data can be integrated to better understand the TRT landscape.

The research question is: How can data from literature and open databases be unified into a structured knowledge graph to identify and visualize key relationships between targets, ligands, radionuclides, and cancer types? The student will design an open, reproducible Neo4j-based knowledge graph combining information from sources such as ChEMBL, UniProt, Open Targets, and PubMed. The initial focus will be on one cancer type (e.g., glioblastoma), with the possibility of extending to ovarian and colorectal cancers if time allows.

Expected outcomes include a modular data integration pipeline, a queryable knowledge graph, and analytical examples demonstrating its utility in hypothesis generation. Successful completion will require skills in Python programming. The knowledge of Neo4j graph databases, data integration, and a general understanding of biomedical terminology and text mining is desirable.