Google DeepMind shows Antigravity finding a novel antimicrobial peptide candidate
In an October 15 technical walkthrough, Google DeepMind described how biologists used the Antigravity agentic science workbench to identify a 97-amino-acid bacteriocin candidate now entering wet-lab validation.
Google DeepMind on October 15, 2026 published a technical walkthrough showing how biologists used Google Antigravity, its agentic development platform, to uncover a novel antimicrobial peptide (AMP) candidate with predicted activity against pathogens involved in respiratory tract infections. The lead is now undergoing wet-lab validation; no measured activity was reported.
Why it matters
DeepMind notes that rising antimicrobial resistance could cause an estimated 10 million deaths a year globally by 2050. Bacteriocins are a promising alternative to traditional antibiotics, but millions of UniProt sequences remain functionally uncharacterized. Google's ProtNLM predicts biological function from raw amino-acid sequences, and Antigravity ties those predictions into an agentic science workbench that compresses fragmented database work into a single session.
The discovery workflow
Researchers set a focused goal: find and computationally validate novel AMPs in Streptococcus pseudopneumoniae. The agent filtered millions of entries to a bacteriocin with UniProt ID A0A0U0D591, 97 amino acids long, from the Lactococcin 972 family.
To check novelty, the researcher steered the agent to BLAST the candidate against curated Swiss-Prot proteins via EMBL-EBI, confirming no meaningful sequence homolog among manually reviewed proteins. Structural analysis pulled the candidate's AlphaFold model and showed its fold closely matches an experimentally determined bacteriocin. Genomic neighborhood analysis found the candidate just upstream of a tightly coupled operon with an ABC transporter permease and another annotated bacteriocin — contextual evidence of a working bacteriocin system.
The agent then drafted a bench protocol covering protein expression, antimicrobial susceptibility testing, target engagement and in vitro toxicity. DeepMind says this candidate and several others from the same pipeline are in wet-lab validation.
Shared 3D canvas
DeepMind also introduced a Biomolecular Structure Viewer inside Antigravity so researcher and agent share a visual state: the agent can highlight functional regions on AlphaFold structures, and scientist annotations flow back for immediate reasoning.
The company frames the walkthrough as lowering the barrier to everyday biology — turning days of database triage into one collaborative session — while stressing that wet-lab results remain the ultimate test.
Source: Google DeepMind Blog.
Related articles
New nuclear waste sensor detects 100 kHz vibrations, could help spot canister cracks
Researchers at the University of Pittsburgh have demonstrated a fiber-optic sensing system capable of detecting...
World’s first fast-spectrum molten chloride reactor experiment nears critical testing
A nuclear experiment backed by the U.S. Department of Energy could deliver crucial operating data...
US Military drones tap 150,000 radios in push to beat battlefield jamming threats
L3Harris Technologies and Skydio are integrating military drones with tactical radio technology to help U.S....
Anthropic grants METR researchers employee-like access for independent AI evaluations
Anthropic said on October 15 it has onboarded six METR employees with company hardware, office space and internal-tool access roughly on par with its own risk teams, and that METR can publish assessments Anthropic cannot veto.
OpenAI launches Rosalind Academic Plan for U.S. university life-science labs
OpenAI on October 15 introduced the Rosalind Academic Plan, offering eligible U.S. university labs sponsored access to GPT-Rosalind and life-sciences research tools under a trusted-access review.