Researcher Uses Codex and ChatGPT to Find Antimicrobials

César de la Fuente's lab is applying Codex and ChatGPT to an urgent global problem: finding new antimicrobials to fight drug-resistant infections. The team is using these AI tools to search both living and extinct genomes for promising antimicrobial candidates, sifting through enormous genomic datasets that would be impractical to review manually. The project highlights how AI coding and reasoning tools can be redirected from software tasks toward computational biology. By pairing large language models with domain-specific workflows, researchers can rapidly generate and test hypotheses about which molecules might have therapeutic potential. Extinct genomes are an especially intriguing source, since they may contain genetic material for defenses that no longer exist in modern organisms. Antibiotic resistance is a growing health crisis. Traditional drug discovery pipelines are slow, expensive, and increasingly failing to keep pace with evolving pathogens. AI-assisted approaches offer a way to compress timelines and surface candidates that conventional methods might overlook. De la Fuente's work is part of a wider trend of using AI agents in the sciences. Rather than replacing researchers, tools like Codex and ChatGPT act as accelerators, handling data-heavy exploration so scientists can focus on validation and clinical relevance. If the approach yields viable candidates, it could become a model for how AI reshapes early-stage drug discovery, turning vast biological archives into a searchable resource for new medicines.