Jure Leskovec (Stanford)

From Data to Discovery: How Foundation Models Transform Biology

Foundation models are revolutionizing biology by learning the underlying structure of complex biological data and transforming how we analyze, interpret, and generate scientific knowledge. In this talk, we explore how these models accelerate discovery by integrating vast datasets, uncovering biological organization, and guiding hypothesis-driven research. We begin with models that learn universal representations of cells across tissues and species. Without explicit supervision, these models capture fundamental biological organization, allowing researchers to analyze, compare, and annotate cells with unprecedented precision. This ability to model biological diversity naturally leads to the next challenge: how to generate and test new biological hypotheses. Large language models, trained on vast scientific literature, can play a crucial role in this process. By leveraging their accumulated knowledge, they help refine hypotheses, automate validation through rigorous testing, and optimize experimental design. These models not only accelerate the scientific process but also provide a scalable, systematic approach to discovery. 

Jure Leskovec is Professor of Computer Science at Stanford University. He is affiliated with the Stanford AI Lab, the Machine Learning Group and the Center for Research on Foundation Models. In the past, he served as a Chief Scientist at Pinterest and was an investigator at Chan Zuckerberg BioHub.  Most recently, he co-founded machine learning startup Kumo.AI. Leskovec pioneered the field of Graph Neural Networks and created PyG, the most widely-used graph neural network library. Research from his group has been used by many countries to fight COVID-19 pandemic, and has been incorporated into products at Facebook, Pinterest, Uber, YouTube, Amazon, and more. His research received several awards including Microsoft Research Faculty Fellowship in 2011, Okawa Research award in 2012, Alfred P. Sloan Fellowship in 2012, Lagrange Prize in 2015, ICDM Research Contributions Award in 2019, and ACM SIGKDD Innovation award in 2023. His research contributions have spanned social networks, data mining and machine learning, and computational biomedicine with the focus on drug discovery. His work has won 12 best paper awards and 5 10-year test of time awards at premier venues in these research areas. Leskovec received his bachelor's degree in computer science from University of Ljubljana, Slovenia, PhD in machine learning from Carnegie Mellon University and postdoctoral training at Cornell University.

About the Seminar Series

The last few years have seen a substantial increase in the reported success of machine learning (ML), and generative artificial intelligence (AI). These impact practices in delivering services from financial institutions to entertainment and medicine. However scientific research also increasingly relies on large data sets, whose analysis leverages ML/AI. This seminar series aims to investigate if and how the paradigm for scientific research has changed or should change to incorporate these new tools and the possibilities they open.

A diverse group of scholars engaged in scientific research, method development, and historical and epistemological investigations will give a 50-minute presentation, followed by discussion.

The event is open to all. Stanford students and postdocs have the opportunity to engage more directly with speakers and topics by enrolling in the Canvas course here.