Case Study: Novartis Institute for Biomedical Research achieves accelerated drug discovery by capturing and connecting biological knowledge with Neo4j

Novartis Captures the Latest Biological Knowledge for Drug Discovery

Novartis, a global healthcare company with thousands of researchers, needed a way to connect decades of sparse experimental data (about a billion historical data points) with terabytes of new phenotypic imaging and the wider medical literature (PubMed’s ~25 million abstracts). The challenge was to let scientists quickly ask complex questions that span genes, diseases and compounds and get evidence‑backed answers that reveal previously hidden associations.

The team built a Neo4j knowledge graph that links genes, diseases and compounds, ingesting PubMed via text mining alongside Novartis’s historical and image data into a model of 15 node types and 90+ relationship types. Graph algorithms locate triangular gene–disease–compound patterns and score association strength, letting researchers rank and drill into correlations. The graph already holds about half a billion relationships (with plans to grow substantially), enabling faster identification of promising compounds and more informed drug‑discovery decisions.

Novartis Institute for Biomedical Research

Stephan Reiling

Senior Scientist