{"slug": "google-explores-biologically-grounded-ai-for-autonomous-wet-labs", "title": "Google explores biologically grounded AI for autonomous wet labs", "summary": "Google Research's VERID framework, a semantic digital twin of a wet lab described in an October 16, 2026 blog post, raised AI-designed aptamer hit rates from 0% (0 of 167 candidates) in round one to 42.1% (131 of 311 tested) by the third iteration against the blood-clotting protein α-thrombin, at roughly $1 to $3 per candidate across about 1,000 candidates per round. The team also used de-novo protein design to produce binders for cancer-related proteins with single-digit nanomolar binding affinities, extending Google's AI-for-science work on the Co-Scientist multi-agent system. The research followed an October 7, 2026 commitment of nearly $1.8 billion to the Virtual Biology Initiative by a coalition including Google DeepMind, Meta, Isomorphic Labs, and US government partners.", "body_md": "# Google explores biologically grounded AI for autonomous wet labs\n\nGoogle Research's VERID framework lets AI agents run real lab experiments, and early results show hit rates climbing from zero to 42.1% in three rounds\n\nAI models are very good at predicting how molecules should behave. Biology is famously bad at taking direction.\n\nGoogle Research is trying to close that gap. In a blog post dated October 16, 2026, the team laid out building blocks for autonomous wet labs, where AI agents design experiments, run them on physical robotics, read the results, and try again.\n\nThe goal is AI that learns from what actually happens in a test tube, not just from what a simulation says should happen.\n\n## Meet VERID, the lab’s digital stunt double\n\nThe centerpiece is a framework called VERID, short for VERified Infrastructure for Discovery Laboratories. [Google](https://cryptobriefing.com/markets/alphabet/) describes it as a semantic digital twin of the lab.\n\nAccording to Google, VERID supports three core functions. Agents can query the system to inspect its state, run time-compressed simulations, and execute protocols safely on physical hardware.\n\nWith those guardrails in place, the agents can run closed-loop cycles. They design a batch of candidates, test them, analyze the outcomes, and refine the next batch.\n\n## From zero hits to 42.1%\n\nThe headline result involves aptamers, short strands of DNA that fold into shapes capable of gripping a target molecule. Google’s agents were tasked with finding aptamers that bind α-thrombin, a protein involved in blood clotting.\n\nThe first round did not go well. Out of 167 candidates tested, none hit, for a success rate of 0%.\n\nBy the third iteration, the picture looked very different. The agents landed 131 hits out of 311 tested, a hit rate of 42.1%.\n\n### AI, tech, and the markets they move—in one daily briefing.\n\nDaily. Free. Join 34,000+ readers across crypto, finance, and policy.\n\nEach round involved approximately 1,000 candidates, at a cost of $1 to $3 per candidate.\n\nThe aptamer work is deliberately cheap. Google frames DNA aptamers as a low-cost, high-throughput sandbox, a place where agents can fail quickly and inexpensively while the methods mature.\n\nThe team then moved to harder targets. Using de-novo protein design, meaning proteins built from scratch rather than adapted from nature, the agents produced binders for cancer-related proteins with single-digit nanomolar binding affinities.\n\nFor non-chemists: binding affinity measures how tightly a molecule holds onto its target, and lower numbers mean a stronger grip. Single-digit nanomolar puts these designs in the tight-binding range that drug developers tend to care about.\n\n## Why simulations alone keep falling short\n\nThe project targets a stubborn problem in computational biology. Predictions made in silico, on a computer, often fail to survive contact with the wet lab.\n\nReal experiments come with measurement errors, contamination, temperature drift, and variables nobody thought to model.\n\nThe work also extends Google’s broader AI-for-science push. It builds on Co-Scientist, the company’s multi-agent system designed to help researchers generate and evaluate hypotheses.\n\n## The money is moving in the same direction\n\nThe research landed shortly after a much larger funding announcement. On October 7, 2026, a coalition including Google DeepMind, [Meta](https://cryptobriefing.com/markets/meta/), Isomorphic Labs, and US government partners committed nearly $1.8 billion to the Virtual Biology Initiative.\n\nThat effort focuses on expanding biological datasets and building virtual cells, computational models meant to simulate how living cells behave.\n\n## What this means for biotech and AI\n\nIf agents can run tight design-test-refine loops at $1 to $3 per candidate and improve hit rates this quickly, the cost of early-stage exploration could fall.\n\nThere are real caveats. The aptamer results come from a sandbox chosen precisely because it is cheap and fast, and the jump from DNA aptamers to complex protein therapeutics is not trivial.\n\nStrong binding affinity is also only one early hurdle. A molecule that grips a cancer protein tightly still has to prove it is safe, stable, and effective in living systems.\n\nGoogle’s emphasis on safe protocol execution acknowledges that letting AI agents control physical lab equipment raises new questions about oversight and error handling.\n\n**Disclosure:** This article was edited by Diego Almada Lopez. For more information on how we create and review content, see our\n\n[Editorial Policy](https://cryptobriefing.com/editorial-policy/).", "url": "https://wpnews.pro/news/google-explores-biologically-grounded-ai-for-autonomous-wet-labs", "canonical_source": "https://cryptobriefing.com/google-verid-ai-autonomous-wet-labs/", "published_at": "2026-10-16 18:35:31+00:00", "updated_at": "2026-10-16 18:53:09.701318+00:00", "lang": "en", "topics": ["artificial-intelligence", "ai-agents", "ai-research", "robotics"], "entities": ["Google", "Google Research", "VERID", "Google DeepMind", "Meta", "Isomorphic Labs", "Virtual Biology Initiative", "Co-Scientist"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/google-explores-biologically-grounded-ai-for-autonomous-wet-labs", "markdown": "https://wpnews.pro/news/google-explores-biologically-grounded-ai-for-autonomous-wet-labs.md", "text": "https://wpnews.pro/news/google-explores-biologically-grounded-ai-for-autonomous-wet-labs.txt", "jsonld": "https://wpnews.pro/news/google-explores-biologically-grounded-ai-for-autonomous-wet-labs.jsonld"}}