DeepMind walkthrough identifies an antimicrobial peptide candidate with Antigravity

Google DeepMind published a technical walkthrough on October 15, 2026 showing biologists using Google Antigravity, its agentic development platform, together with ProtNLM, a model that predicts protein function from amino acid sequences, to scan millions of previously uncharacterized UniProt entries. The session identified a novel antimicrobial peptide candidate with predicted activity against pathogens involved in respiratory tract infections.

Sourcing note: this report rests on a single reporting chain, DeepMind's own walkthrough and summaries of it. The candidate is a computational lead, not a validated drug, and DeepMind says it is now undergoing wet lab validation. This item will be updated when validation results or independent reporting appear.

Sources: Google DeepMind; Unite.AI