Abstract: We are laying the foundations of AI for molecular discovery and drug design, eventually enabling AI to learn and innovate on its own. Instead of training separate models for every task, we leverage large language and geometric models across many tasks through fine-tuning and few-shot prompting. Central to our approach is the integration of molecular structures, biological knowledge, and genomic data into AI models. We are advancing self-supervised learning to leverage multi-omics datasets and geometric deep learning to model the geometry of molecules. I describe PINNACLE AI, contextual AI models for single-cell protein biology. PINNACLE models enhance 3D structures of protein-protein interactions in immune-oncology, predict drug effects across cell types and cell states, and nominate therapeutic targets in a cell-type-specific manner. For drugs to be effective, they must act on biological targets in core disease processes. I describe our multimodal sequence-structure generative models that design molecules to serve as optimal binders with desired biochemical properties. Finally, candidate drugs need to be matched to patient benefits. I introduce TxGNN, a knowledge graph AI model for zero-shot prediction of therapeutic use across over 17,000 diseases, enabling drug repurposing for 7,000 rare diseases, with a mere 5% having FDA-approved drugs. TxGNN’s predictions align with medication use across millions of medical records.
Bio: Marinka Zitnik (https://zitniklab.hms.harvard.edu) is an Assistant Professor of Biomedical Informatics at Harvard Medical School, at Kempner Institute for the Study of Natural and Artificial Intelligence at Harvard University and at Broad Institute of MIT and Harvard. Zitnik investigates foundations of AI that contribute to the scientific understanding of medicine and therapeutic design, eventually enabling AI to learn and innovate on its own. Her research won best paper and research awards, including the Kavli Fellowship of the National Academy of Sciences, Kaneb Fellowship award at Harvard Medical School, NSF CAREER Award, awards from the International Society for Computational Biology, International Conference in Machine Learning, Bayer Early Excellence in Science, Amazon Faculty Research, Google Faculty Research, Roche Alliance with Distinguished Scientists, and Sanofi iDEA-iTECH Award. Zitnik founded Therapeutics Data Commons, a global open-science initiative to access and evaluate AI across stages of development and therapeutic modalities, and she served as the faculty lead of the AI4Science initiative.