AgentPlays
Google DeepMind says Antigravity helped find a novel antimicrobial peptide candidate
On Oct 15, Google DeepMind published a technical walkthrough of biologists using Google Antigravity, its agentic development platform, to hunt for antimicrobial peptides. In one session the agent filtered UniProt with ProtNLM, checked novelty against Swiss-Prot, compared the AlphaFold structure, and inspected the genomic neighborhood around a 97-amino-acid Lactococcin 972-family bacteriocin (UniProt A0A0U0D591) from Streptococcus pseudopneumoniae. DeepMind says the candidate has predicted activity against pathogens tied to respiratory tract infections and is now in wet-lab validation, alongside others from the same pipeline. The post also introduces a Biomolecular Structure Viewer inside Antigravity so the researcher and agent share a 3D canvas.
How you’d use it: If you work on antimicrobial discovery, open the DeepMind walkthrough and try the same UniProt → ProtNLM → AlphaFold → genomic-context path inside Antigravity before you design a wet-lab assay.