One signal

Update

What changed

Google DeepMind’s fuller account adds a concrete example of its Antigravity chemistry workflow, plus a free access offer for 10,000 scientists. The team says the agent sifted through thousands of compounds and proposed rigidified macrocycles with synthetic routes in about an hour.

The post describes a workflow built around Google Antigravity’s new Cheminformatics Skills, with RDKit and connections to ChEMBL and PubChem. In its lung-cancer-treatment example, the agent extracts compounds and data from literature, performs matched molecular pair analysis, builds potency models and suggests new designs.

The company says the cheminformatics skills are available in Antigravity today.

DeepMind is clear about the important boundary: predicted designs are hypotheses, not finished drugs, and every proposed compound still needs to be synthesised and tested experimentally.

Sources and evidence

• Agentic Cheminformatics with Google Antigravity – Google DeepMind: Google DeepMind says its Antigravity Cheminformatics Skills workflow used an agent to analyse thousands of compounds and propose rigidified macrocycles with synthetic routes in about an hour; the skills are available today, and the company is offering six months of its highest usage tier to 10,000 scientists through an application process.

Independent WittyWires Watcher; not an official account or feed.