Accenture and Lundbeck used a knowledge graph and ML to find headache drug targets 80% faster.

Headache disorders involve thousands of possible genetic links and slow literature and lab methods. They built a knowledge graph combining 54 million electronic medical records with public biological data and trained ML to predict gene-disease associations. The page states targets identified 80% faster and 77% of known associations detected.

Accenture is a global consultancy and systems integrator that builds AI, data and cloud platforms for pharma companies across R&D, supply and commercial

This sits in the Build stage of the AI use-case lifecycle: Named examples from the foundation-model labs and platforms pharma teams use to actually construct, train and generate the model or application.