Artificial intelligence holds tremendous promise in enabling scientific breakthroughs and discoveries in diverse areas. We investigate applied machine learning with a current focus on networked systems that require infusing structure and knowledge.
Our research strategy is to create foundational models, including pre-trained, self-supervised, multi-purpose, and multi-modal models trained on broad networked data at scale. This entails identifying ways to infuse knowledge and structure into models to address uncertainty and enable broad generalization, and producing actionable and trustworthy representations that advance the biological problem past the state of the art and open up new opportunities.
The state of a person is described with increasing precision incorporating modalities like genetic code, behaviors, therapeutics, and the environment—the challenge is how to reason over these data to improve decision making. Our research creates new avenues for accelerating the development of therapeutics, fusing biomedical knowledge and patient data, and giving the right patient the right treatment at the right time to have medicinal effects that are consistent from person to person and with results in the laboratory.
For centuries, the method of discovery—the fundamental practice of science that scientists use to explain the natural world systematically and logically—has remained largely the same. However, the natural world is interconnected, from all facets of genome regulation to the molecular level and to the population level. These interactions at different levels give rise to a bewildering degree of complexity. Our research disentangles this complexity and develops artificial intelligence tools to guide discovery in biomedical sciences and produce interpretable outputs that lend themselves to actionable hypotheses.