Google DeepMind uses Antigravity platform to speed antimicrobial peptide discovery

Google DeepMind researchers Pengfei Tian, Maxwell L. Bileschi, Zoi Tsangalidou, Sukhdeep Singh, and Lucy J. Colwell have used Google Antigravity to accelerate the search for novel antimicrobial peptides. DeepMind Research News and Google DeepMind both reported that the tool addresses rising levels of antimicrobial resistance by speeding up the search for new therapeutic candidates. [1][2]

DeepMind Research News reported that Google Antigravity operates as an agentic science workbench for biologists. The platform enables scientists to discover, structurally evaluate, and prioritize therapeutic targets within hours, consolidating what previously required fragmented searches across multiple databases into a single session. [1]

Google Antigravity is a software development platform where artificial intelligence agents write, run, and test code autonomously. Google released the platform on November 18, 2025, alongside its Gemini 3 model as an integrated development environment. In May 2026, Google added Antigravity 2.0 as a standalone desktop application featuring a command-line interface and a software development kit. [4]

The platform supports various artificial intelligence models, including Gemini 3.1 Pro, Gemini 3 Flash, Anthropic Claude Sonnet and Opus 4.6, and gpt-oss-120b. Direct application uses of Gemini models through systems like Antigravity offer raw model access without the real-time web access or additional safety layers present in standard web interfaces. [5]