High-throughput methods leveraging robotics and computer vision for the development of therapeutic phage cocktails.

Abstract

We present the high-throughput automated screening techniques that are being used to develop bacteriophage-based therapeutic products currently under investigation in human clinical trials to combat urinary tract infections¹. By integrating modern liquid handling robotics, standardized phenotypic assays, and computer vision-based enumeration, we established a platform capable of reproducibly screening large collections of phages against clinically derived bacterial strain panels. This approach enabled systematic assessment of phage-bacteria interactions at scale, facilitating the identification and optimization of phage cocktails with broad in vitro activity. Although bacteriophage therapy has long been investigated as a strategy for treating bacterial infections, few frameworks exist for developing phage combinations in a reproducible and scalable manner. The methods outlined here address this gap and aim to support the broader development of therapeutic assets available to combat antibiotic resistance.