OpenAI and partners used thousands of AI agents to scan millions of mouse brain cells in 32 minutes. The leads are early, but they could speed up disease research.
OpenAI says thousands of AI agents working at once analyzed experimental data from millions of mouse brain cells in 32 minutes. The team estimates that doing the same review by hand could take months to a year. The work turned up possible clues about brain diseases, but it is not a cure or a treatment.
OpenAI worked with PerturbAI and Scripps Research on the study, announced Wednesday. The data showed what happens to brain cells when different genes are disrupted. Instead of one AI model doing everything, each agent took one genetic change, studied how the cells responded and searched scientific papers for explanations. The results were pooled into what the researchers call a "reasoning atlas." It links genes, cell changes and existing research.
The agents ran on GPT-Rosalind, OpenAI's specialized AI model for life sciences.
One lead involves a gene called DCTN1, which is tied to Perry syndrome. That is a rare brain-degeneration condition that can cause Parkinson-like symptoms. When researchers disrupted the gene, genes that make cholesterol became more active. The agents linked that to earlier research suggesting that trouble moving cholesterol around inside cells can trigger extra cholesterol production.
The agents also noticed that two genes, Tsc1 and Tsc2, behaved differently. The genes work together to regulate cell growth and metabolism. Yet disrupting each one had different effects in certain neurons in the cerebellum, the part of the brain that helps coordinate movement.
The researchers also tested whether the crowd approach beat a single powerful AI system. It did on two measures:
These are research leads, not proven results. Scientists still need to test them in the lab. The data came from mouse cells, and nothing here shows a new treatment for Alzheimer's or Parkinson's disease.
The larger point is about speed. Experiments can produce data on millions of cells, but researchers rarely have time to chase every surprising result. Tomasz Nowakowski, a neurological surgery professor at the University of California, San Francisco, said in OpenAI's announcement that useful clues often sit scattered across different fields. Connecting them can open new ways of thinking about disease.
AI agents helped surface early brain-disease leads in 32 minutes, but those leads still need lab testing before they matter to patients.
A tech writer fed his Windows error logs to ChatGPT and learned his encrypted drive had no recovery key, a problem that could have locked him out for good.
Read more: ChatGPT read his laptop's error logs and found three hidden problemsA TikTok Shop affiliate says a strong first line, an honest demo and steady posting earned him six figures last year. Here is how it works and what it costs to try.
Read more: How a TikTok creator earns six figures with a 3-second opening lineGoogle's Gemini can now edit a PDF inside the chat, so fixing a travel plan no longer means downloading the file and starting over in another app.
Read more: Gemini can now edit your PDF itinerary without starting over
