Google DeepMind uses Antigravity AI to identify potential new antimicrobial peptide

Researchers identified the candidate in a single AI-assisted session, with laboratory testing now underway to assess its predicted activity against pathogens linked to respiratory infections

Google DeepMind researchers have used Antigravity to identify a potential antimicrobial peptide, which is now undergoing laboratory validation

Google DeepMind researchers have used Google Antigravity, the company's AI-powered development platform, to identify a previously uncharacterized antimicrobial peptide candidate, demonstrating how AI agents can support scientific discovery by connecting multiple research tools within a single workflow.

The candidate, which has predicted activity against pathogens associated with respiratory tract infections, was identified during an interactive session in which a researcher directed the scientific investigation while Antigravity handled database searches, protein analysis and the review of existing literature.

The discovery remains at the computational stage. The peptide, alongside several other candidates identified through the same process, is now undergoing wet-lab validation to establish whether the predictions hold up experimentally.

Google DeepMind has been extending Antigravity beyond its original coding applications, combining its Gemini models with specialist scientific tools, including AlphaFold and AlphaGenome.

Pushmeet Kohli, Chief Scientist at Google Cloud and VP Science at Google DeepMind, highlighted the speed of the process in a LinkedIn post:

"As an example, in a single session using Antigravity, our scientists working on Antimicrobial resistance surfaced a novel antimicrobial peptide candidate."

How Antigravity narrowed millions of proteins to a potential candidate

The research focused on antimicrobial peptides, naturally occurring molecules that can act against bacteria and are being investigated as potential alternatives to conventional antibiotics.

Finding suitable candidates involves searching enormous biological databases, many of which contain protein sequences whose functions have never been established.

Google DeepMind used ProtNLM, its Protein Natural Language Model, to help address this problem. The model examines amino acid sequences and predicts their biological functions, providing information about proteins that have not previously been characterized.

In the demonstrated workflow, researchers tasked Antigravity with investigating potential antimicrobial peptides in Streptococcus pseudopneumoniae.

The system narrowed millions of database entries to a small group of candidates, identifying a 97-amino-acid peptide belonging to the Lactococcin 972 family.

Further analysis involved comparing the candidate against reviewed protein sequences, examining its predicted three-dimensional structure using AlphaFold data and investigating neighboring genes for supporting biological evidence.

Although the analysis found no meaningful sequence match among manually reviewed proteins, structural comparisons indicated similarities with an experimentally determined bacteriocin, a type of antimicrobial peptide produced by bacteria.

Antigravity also reviewed existing literature to develop hypotheses about how the molecule might work and which pathogens it could potentially affect.

These findings provide computational evidence for further investigation, but do not establish that the candidate is an effective antimicrobial treatment.

New 3D viewer and academic credits extend the research tools

Alongside the case study, Google DeepMind has introduced a Biomolecular Structure Viewer within Antigravity, allowing researchers and AI agents to examine and interact with three-dimensional protein structures in a shared workspace.

The viewer supports structures retrieved from the AlphaFold Protein Structure Database or researchers' own predictions. Scientists can highlight specific regions, add notes and ask the agent questions about particular molecular features, with those interactions shared directly between the researcher and the AI system.

The company has also introduced an Antigravity for Science credits program to support academic researchers using the tools, although the supplied announcement does not specify the value of the credits or detailed eligibility requirements.

For the antimicrobial peptide project, the next stage is experimental validation. Researchers are testing the identified candidate and others generated through the same workflow to determine whether their predicted biological activity can be demonstrated in the laboratory.

ETIH Innovation Awards 2027 open for entries

The ETIH Innovation Awards 2027 are now open for entries, with 26 categories recognising companies, institutions, projects and individuals making a measurable impact across education technology. The global awards span AI in education, digital learning, higher education, skills and workforce, inclusion, education operations, evidence and impact, with entries managed through ETIH’s new online awards platform.

Entries are open worldwide, with the early-bird deadline on Friday 18 December 2026 and the final entry deadline on Friday 26 February 2027 at 23:59 UK time. 

Explore the 2027 categories and judging criteria, read the How to Enter guide or start an entry through the ETIH Innovation Awards 2027 platform.