Accelerating the search for antimicrobial peptides with Google Antigravity
Google DeepMind describes how its Antigravity platform helped biologists search for new antimicrobial peptides, combining UniProt annotation, sequence and structure comparisons and literature review in one session. It surfaced a bacteriocin candidate from Streptococcus pseudopneumoniae with predicted activity against respiratory pathogens, now in wet-lab testing.
What the article says
Antimicrobial resistance is a global health threat, with an estimated 10 million deaths a year possible by 2050. Antimicrobial peptides, especially bacteriocins, are a promising alternative to conventional antibiotics, but many candidates sit in the uncharacterized dark matter of protein databases: UniProt alone holds millions of sequences of unknown function. Google DeepMind presents a method in which Google Antigravity acts as an agentic science workbench to discover, structurally evaluate and prioritize candidate proteins in a single session.
The approach relies on ProtNLM, a model that predicts a protein's function from its sequence, and on an embedded 3D viewer shared by the researcher and the agent. In the example, from the bacterium Streptococcus pseudopneumoniae, the agent selected a 97-amino-acid bacteriocin with no meaningful homolog in Swiss-Prot, whose predicted structure matches a known bacteriocin, and drafted an experimental plan. The candidate is in wet-lab validation along with others; the page presents it as a biologically plausible lead, not yet validated.
Read the official article AnnuIA profile
Summary written by AnnuIA in its own words from the official page of Google DeepMind. The content and rights belong to the publisher: please read the original article.