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. Working at the forefront of this development, I seek to harness computational power to find molecules with antibacterial activity. I use synthetic biology and computational tools to determine features contributing to this activity and train computers to find— or design— candidate molecules and tweak their structures virtually. Experimentation is reserved for validating computer predictions, saving time, labor, and expense. With machine-based molecular discovery, I explore proteins and peptides as engineering scaffolds. My approaches diversify proteins, such as host defense peptides (HDPs), beyond their natural variation. For example, to increase their antimicrobial properties, we trained a computer to execute a fitness function that selects for structures that interact with bacterial membranes, thereby converting several HDPs into the first artificial antimicrobials that kill bacteria both in vitro and in animals. By investigating these exciting possibilities, I aim to build machine-made antibiotics to combat infectious diseases and develop clinical applications for autonomously generated synthetic molecules. Computer-made drugs may help to replenish our arsenal of effective drugs and outpace the evolution of antibiotic resistance.