Charlotte Bunne

Wednesday, December 10, 2025

4:00 - 5:00 pm (refreshments at 3:30 pm)

Monadnock (Broad Institute - 415 Main St., room 2040)* and virtually at broad.io/ewsc

*If you do not have a Broad badge, please arrive 10 minutes early with a different ID to register with security and be escorted up to Monadnock.

Add to calendar

Abstract:

To realize the promise of precision oncology, we need AI systems that can move beyond diagnosis and analysis to become predictive simulators of disease and therapy. In this talk, I will describe our recent work on innovating new AI architectures that combine multi-modal foundation models with generative modeling for treatment effect prediction.

We design foundation models to capture the complexity of cancer by learning across biological scales, from molecular interactions to tissue architecture. Architecturally, our approach builds on principles from multi-modal representation learning and large-scale vision models, enabling the integration of proteomics, pathology, and clinical annotations into a unified virtual patient space. Through unsupervised learning and multi-scale neural network design tailored to high-dimensional multiplexed imaging, we create representations where new biopsy samples can be consistently mapped and biomarkers can be discovered, supporting downstream analyses of molecular, morphological, and spatial complexity. On top of these representations, we extend the framework with generative modeling to move from analysis to simulation, predicting therapeutic responses, resistance dynamics, and enabling in silico exploration of treatment outcomes.

Our broader vision is to develop Virtual Patient Labs as digital counterparts of patients, where integrative models and generative simulation converge to anticipate disease trajectories and treatment responses. These environments would make in silico experimentation possible at patient scale and open new avenues for AI-guided decision support in oncology.

Biography:

Charlotte Bunne is an Assistant Professor at EPFL, jointly appointed in the School of Computer and Communication Sciences (IC) and the School of Life Sciences (SV). She is a member of the EPFL AI Center and the Swiss Institute for Experimental Cancer Research (ISREC), and is affiliated with the Precision Oncology Unit at the University Hospital in Geneva. She previously held postdoctoral positions at Genentech and Stanford, working with Aviv Regev and Jure Leskovec, after completing her PhD in Computer Science at ETH Zurich under the supervision of Andreas Krause and Marco Cuturi. During her doctoral studies, she was a visiting researcher at the Broad Institute of MIT and Harvard and a visiting student at MIT. Her honors include the AI2050 Early Career Fellowship, a Fellowship of the German National Academic Foundation, and two ETH Zurich Medals.