abstract
- New antibiotic combinations are needed to improve the treatment of tuberculosis. Larkins-Ford and colleagues share a framework that combines in vitro pairwise drug response data and machine learning to rationally prioritize combinations for clinical development.1.