Google Antigravity is an agentic science workbench for antimicrobial peptide discovery. It compresses fragmented database searches into one session. ProtNLM predicts protein function from amino acid sequences in UniProt. Researchers used it to find a novel peptide candidate against respiratory pathogens. The candidate is now in wet-lab validation. A shared 3D Biomolecular Structure Viewer enables interactive collaboration. The agent retrieves structures, highlights regions, and reasons over user annotations. The workflow replaces custom scripts and disconnected tools. It narrows candidates to high-confidence leads before costly lab tests. The same pipeline can target other pathogens or sequence lengths.
What changed
Biologists queried databases, exported files, and used disconnected tools for weeks to find candidates.
What it unlocks
Discovering, evaluating, and prioritizing novel antimicrobial peptides in a single collaborative session.
- 10 million annual deaths by 2050 from antimicrobial resistance
- millions of uncharacterized UniProt proteins
- 97 amino acids in candidate bacteriocin
What you need to act on it
- access to Google Antigravity
- UniProt database access
- AlphaFold Protein Structure Database access
Sources