January 17, 2014 - 10:30am
Wright Rieman, Rm. 260
"A Novel Intersection of Computation, Chemistry, and Biology in the Study of Tuberculosis"
Spring 2014 Colloquium Series
Joel S. Freundlich
|Abstract: Tuberculosis represents a global health pandemic, claiming the lives of 1.4 million people per annum. The continued rise of drug resistance and the need for shorter therapeutic regimens consistent with antiviral cocktail therapy are driving the study of the basic biology of the causative pathogen -Mycobacterium tuberculosis. In an effort to better understand how M. tuberculosis persists in the face of the host immune response, we have sought a novel combination of computational, chemical, and biological tools. This seminar will detail our efforts to leverage a family of small molecules to probe M. tuberculosis fatty acid biosynthesis, which is critical to maintenance of the cell wall. Critically, this approach began with an empirical high-throughput screen and we pursued only 1 active molecule and essentially discarded the other 49,999 compounds assayed. We hypothesize that the entire data set (actives and inactives) in a whole-cell high-throughput phenotypic screen can be leveraged in chemical probe and drug discovery. We demonstrate the construction and prospective validation of machine learning models for antitubercular whole-cell efficacy and lack of relative model mammalian cell cytotoxicity. Single and novel dual-event models have been harnessed to offer hit rates significantly in excess of empirical screening methods and have identified a set of promising chemical probes to explore M. tuberculosis infection biology while seeding potential drug discovery leads.|
|Host: Professor Lawrence Williams|
|~ Coffee/tea will be served prior to lecture ~|