Researchers at the University of Pennsylvania introduced ApexGO, an AI method designed to improve peptide antibiotic candidates rather than relying on broad screening of very large molecular libraries. ApexGO starts with a small set of candidates and iteratively selects modifications using a predictive algorithm to guide optimization. The team reported lab validation where 85% of ApexGO-generated molecules halted bacterial growth and 72% outperformed the parent peptides. In mice, two optimized antimicrobial peptides reduced bacterial counts to levels comparable to polymyxin B, the group said. The study positions AI as a more directed “optimization” tool for antimicrobial peptide engineering, potentially compressing timelines from initial hits to functional molecules in the context of rising antibiotic resistance.