Eureka! GSK's Kim Branson on driving innovation with AI and machine learning
Mar 11, 2025
Kim Branson, Senior VP and Global Head of AI and Machine Learning at GSK, shares her fascinating career journey from structural biology to leading AI initiatives. She delves into how GSK is utilizing machine learning to enhance drug development, focusing on precision medicine in oncology. Kim discusses the importance of combining ML experts with domain specialists and treating data as a strategic asset. She also highlights the future of multimodal datasets from wearables and the significance of responsible AI governance in pharmaceutical innovation.
AI Snips
Chapters
Transcript
Episode notes
From Crystallography To AI In Pharma
- Kim Branson described moving from structural biology and x-ray crystallography into machine learning applied to drug design in the late 90s and early 2000s.
- She recounted roles at Vertex, work with early Google folks, and transitions through Genentech to GSK where she has led AI/ML for five years.
Models As Biomarkers Linking Data To Outcomes
- GSK applies ML across genetics, cellular and clinical imaging, and active learning to predict treatment effects and discover disease heterogeneity.
- Branson views biomarkers as models that link multimodal data to outcomes and support precision prescribing and software-enabled assets.
Instrumented Trials Revealed Predictive Signals
- Kim used the Repraversin (Bepi) hepatitis program as an example where heavily instrumented phase‑2 trials produced blood and antigen biomarker data.
- She explained how computational pathology plus outcome data let ML predict response directly from slides and spatial cellular context.