Verge Labs Uses AI and 12,000 Brain Scans to Hunt for New CNS Drugs
A rebranded AI lab is training disease models on massive brain datasets to speed up drug discovery for neurological conditions.
Summary
Verge Labs, formerly Verge Genomics, has relaunched as an AI-focused drug discovery company targeting brain and nervous system diseases. The company is training large AI models on VergeDB, a proprietary dataset containing over 12,000 brain transcriptomes, 15 million single-cell profiles, and matched protein, genetic, and clinical data from roughly 6,000 patients. A standout feature is a virtual biopsy capability that can extract brain-relevant biological signals from a simple blood draw. Their platform has shown 83 percent accuracy in predicting valid drug targets, and their first AI-identified drug candidate completed a Phase 1b clinical trial in late 2025. The company is now partnering with drug developers focused on central nervous system disorders and has brought in senior talent from leading biotech and AI organizations.
Detailed Summary
Verge Labs represents a significant moment in AI-driven drug discovery, particularly for brain health — one of the most underfunded and difficult areas in medicine. Neurological diseases like ALS, Parkinson's, and Alzheimer's remain largely without disease-modifying treatments, and the company is positioning its AI platform as a way to dramatically accelerate the identification of effective drug targets.
The company's core asset is VergeDB, a multimodal human brain dataset built from over 12,000 brain transcriptomes drawn from approximately 6,000 patients. It also includes 15 million single-cell profiles alongside proteomic, genomic, and clinical data. Critically, the database incorporates linked blood, cerebrospinal fluid, and imaging data from living patients, enabling what the company calls a virtual biopsy — the ability to infer brain biology from a routine blood draw rather than invasive tissue sampling.
Verge's AI platform has achieved an 83 percent validation rate for drug targets it surfaces, confirmed against experimental data across a decade of internal programs. That figure is notably high in an industry where target identification remains a major bottleneck. The company's first AI-discovered drug candidate completed a Phase 1b clinical trial in late 2025, with those trial results now feeding back into VergeDB as a training signal — creating a self-improving research loop.
For longevity-focused readers, this matters because neurodegeneration is one of the leading causes of age-related disability and death. Better tools for identifying drug targets and enabling early biological monitoring of brain health could translate into earlier interventions and more effective therapies in the coming decade.
Caveats apply: the 83 percent validation rate is an internal company metric and has not yet been independently peer-reviewed. The Phase 1b trial has not published full results, and commercial partnerships are still in early stages. Independent scrutiny of platform performance will be essential before drawing firm conclusions.
Key Findings
- VergeDB holds 12,000+ brain transcriptomes and 15 million single-cell profiles from ~6,000 patients.
- A virtual biopsy tool can infer brain biology from a standard blood draw, avoiding invasive procedures.
- The AI platform validated drug targets at 83% accuracy across ten years of internal programs.
- First AI-discovered drug candidate completed Phase 1b trial in late 2025, results fed back into training data.
- Senior hires from Altos Labs and Calico signal serious longevity-sector credibility and ambition.
Methodology
This is a news report based on a company press announcement from Longevity.Technology, a credible industry publication. Evidence is entirely company-reported and has not been independently peer-reviewed. No primary research paper accompanies the launch.
Study Limitations
The 83 percent target validation rate is self-reported by Verge Labs and lacks independent peer-reviewed confirmation. Phase 1b trial data has not been publicly released or published. Claims about platform performance should be verified against future independent studies.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.