From the source
Google DeepMind's technical walkthrough shows how biologists used
Google Antigravity, its agentic development platform, to find a novel antimicrobial peptide candidate with predicted activity against respiratory tract infection pathogens.
- The lead is now undergoing wet-lab validation, the post says.
- Antigravity provides an agentic science workbench that compresses a fragmented search across multiple databases into a single session, and it uses ProtNLM, a deep learning model that predicts biological function directly from raw amino acid sequences, to scan millions of uncharacterized UniProt proteins.
- In the documented session, the agent filtered millions of entries down to a bacteriocin of 97 amino acids from the Lactococcin 972 family, and the researcher steered checks against Swiss-Prot via NCBI-BLAST, a 3D model from the AlphaFold Protein Structure Database, and the candidate's genomic neighborhood.
- The flanking genes placed the candidate 6 base pairs upstream of a tightly coupled operon with an ABC transporter permease and ATP-binding protein, which the post calls strong contextual evidence of a working bacteriocin system.
