Algorithms for Biomolecular Structures at Proteome Scale

Breakthroughs in deep learning methods for protein structure prediction have transformed structural biology. At the same time, cryo-electron microscopy (cryo-EM) has enabled the reconstruction of molecular structures and the capture of movies with unprecedented detail. As machine learning continues to transform structural biology, what are the next frontiers for biomolecular structure determination?

In this Presidential Lecture, Ellen Zhong will describe the algorithmic challenges at the frontier of structure determination via cryo-EM. She will provide an overview of cryoDRGN, a machine learning system for heterogeneous cryo-EM and cryo-ET reconstruction. Along the way, she will overview recent progress her group has made in reconstructing complex mixtures, developing challenging benchmarks for structural heterogeneity and visualizing dynamic biomolecular complexes inside the cell. Finally, she will discuss how multimodal foundation models that integrate sequence, structure and imaging data can enable new approaches to reconstructing dynamic biomolecular complexes at scale, pointing toward a data-driven paradigm for visual proteomics.

Zhong is an assistant professor of computer science at Princeton University. Her group’s research spans methodological research in AI and computer vision, as well as close collaboration with experimentalists in molecular and structural biology. She previously worked on the AlphaFold team at Google DeepMind and on molecular dynamics for drug discovery at D. E. Shaw Research. Her work has been recognized with the NIH Director’s New Innovator Award, the Schmidt Sciences AI2050 Early Career Fellowship and a Major Society Award from the Microscopy Society of America. She earned her Ph.D. from MIT in 2022.