We believe that artificial intelligence is going to define the future of drug discovery. We leverage the state-of-the-art in high-throughput biology and contemporary artificial intelligence approaches to discover and mechanistically investigate novel antibiotics. Specifically, we develop and implement machine learning methods to predict structurally novel antibiotics from vast in silico databases and de novo design structurally novel antibiotics from scratch. We are also building models to help us elucidate the biochemical mechanisms of action and pharmacological properties of new antibiotics. Our goal is to increase the rate and decrease the cost of getting new antibiotics to patients.

An ambitious goal in antibiotic discovery is inventing novel therapies that target biological processes that modulate infection in vivo. However, a comprehensive list of “in vivo essential” biological processes, and a holistic understanding of how bacterial cell physiology and host physiology are interrelated, are currently lacking. We investigate bacterial physiology and host tissue physiology using multi-omics approaches to (a) understand how pathogen and host cellular responses change as a function of infection progression; and (b) invent functionally novel antibacterial therapies that modulate the pathogen and/or the host to efficiently suppress the severity of infection.