OpenAI has published a case study highlighting how researcher César de la Fuente uses its Codex and ChatGPT tools to speed up the discovery of new antimicrobial molecules. His lab mines large genomic datasets from both living organisms and extinct species to identify peptides capable of neutralizing bacteria that have become resistant to existing treatments.
Antimicrobial resistance is widely regarded by health authorities as one of the major threats to global public health, as conventional treatments gradually lose effectiveness against evolving pathogens. In response, de la Fuente's team is turning to AI to search molecular spaces too vast to screen manually.
In this workflow, Codex is used to generate and automate bioinformatics analysis code, allowing large volumes of genetic sequences to be processed quickly. ChatGPT is used to formulate hypotheses, interpret intermediate results, and guide subsequent research steps, in an iterative loop between automated computation and assisted reasoning.
The case study, published directly by OpenAI, reflects a broader trend of using generative assistants as productivity tools in biomedical research, complementing rather than replacing laboratory validation. It is worth noting that such company-authored case studies also serve as marketing for OpenAI's own products, and that the real-world efficacy of any identified molecules will ultimately depend on preclinical and clinical trials still to come.