Google uses Antigravity agent to find a candidate antimicrobial peptide

Google researchers used its Antigravity agentic development platform and ProtNLM model to identify a novel antimicrobial peptide candidate with predicted activity against pathogens in respiratory tract infections.

The candidate is a computational lead that is now undergoing wet-lab validation. The workflow runs in a single session, combining UniProt searches, AlphaFold structure analysis and a 3D structure viewer.