Two independent Nature Biotechnology papers detail how deep‑learning approaches accelerated antibacterial discovery. One study combined phenotypic screening with deep neural networks to prioritize compounds with antibacterial activity; the other trained models on high‑throughput screening data to identify novel scaffolds that evade known resistance mechanisms. Both teams validated hits in in vitro bacterial assays and highlighted the methods’ ability to surface structurally novel chemotypes. Authors and peer reviewers framed the work as a step toward data‑driven antibacterial pipelines that can complement traditional medicinal chemistry and phenotypic campaigns.