AI uncovers new antibiotic candidates for drug-resistant gonorrhea

Machine Learning Uncovers New Antibiotic Candidates Against Drug-Resistant Gonorrhea

Researchers at the Wyss Institute for Biologically Inspired Engineering at Harvard University have used deep learning to identify new antibiotic compounds effective against multi-drug resistant Neisseria gonorrhoeae, according to a study published Wednesday in Science Translational Medicine.science

The work arrives amid growing alarm over gonorrhea’s resistance to existing treatments. The World Health Organization lists N. gonorrhoeae as a high-priority pathogen, and in 2023, Massachusetts reported the first U.S. case of gonorrhea showing reduced response to five classes of antibiotics.mass

A Computational Search Through Millions of Compounds

The team employed directed message-passing neural networks to screen large chemical libraries for molecules with antigonococcal activity, identifying candidates that are structurally distinct from existing antibiotics. The approach builds on a series of deep learning-driven antibiotic discoveries at the Wyss Institute and MIT, including earlier work that identified compounds effective against MRSA published in Nature in 2023 and a generative AI framework published in Cell that yielded compounds NG1 and DN1 against gonorrhea and MRSA respectively.nih

Unlike those earlier efforts, the new study validated its lead compounds using the Wyss Institute’s microfluidic Organ Chip technology—specifically a human vagina chip that replicates the vaginal tissue microenvironment—in addition to a mouse vaginal infection model.harvard

Validation Beyond the Petri Dish

The use of the organ chip platform represents a methodological advance in antibiotic preclinical testing, allowing researchers to assess drug efficacy in a system that mimics human physiology more closely than standard cell cultures. The Wyss Institute has previously used its organ chip technology to identify drug candidates for COVID-19 and bacterial vaginosis.harvard

The findings add to a wave of AI-driven antibiotic candidates reaching preclinical validation stages. Two new antibiotics for gonorrhea—gepotidacin and zoliflodacin—were approved by the FDA in late 2025, marking the first new drugs for the disease in decades. Whether the newly identified compounds will advance to clinical trials remains to be determined, but the integration of deep learning with organ-on-chip validation offers a faster pipeline for identifying and testing candidates against resistant pathogens.harvard