Abstract

Until now, the natural world has supplied us with antibiotics. Bacteria, however, are increasingly resistant to these drugs. The next generation of antibiotics will likely come not from nature but from computer-based discovery. Computer-driven approaches have the potential to outperform humans, as demonstrated for pattern recognition of images and text. In order for machines to discover novel drugs and optimize antimicrobial properties, they have to be able to understand, read and write molecules. In this talk, I will describe our efforts in developing computational approaches for antibiotic discovery. I will discuss how we trained a computer to execute a fitness function following Darwin’s algorithm of evolution to select for structures that interact with bacterial membranes, yielding the first artificial antimicrobials that kill bacteria both in vitro and in animal models. My lab has also developed pattern recognition algorithms to mine the human proteome, identifying throughout the body thousands of antibiotics encoded in proteins with unrelated biological function. Computer-made drugs may help to replenish our arsenal of effective drugs.