Machine-learning approaches prevent post-treatment resistance-gaining bacterial recurrences

Abstract

Despite susceptibility testing, recurrent infections are common and are associated with resistance. Using whole-genome sequencing, Stracy et al. demonstrated that recurrence is often driven by a different strain than the original infection. By machine-learning analysis, they developed an algorithm for patient-specific recommendations to minimize antimicrobial resistance (AMR) at the individual-patient level.

Keywords: antimicrobial resistance; infection recurrences; machine-learning; whole-genome sequencing.