{"version":1,"type":"story","url":"https://digestai.news/story/openai-agents-study-parkinson-like-disease-data-in-32-minutes","json":"https://digestai.news/story/openai-agents-study-parkinson-like-disease-data-in-32-minutes.json","markdown":"https://digestai.news/story/openai-agents-study-parkinson-like-disease-data-in-32-minutes.md","slug":"openai-agents-study-parkinson-like-disease-data-in-32-minutes","headline":"OpenAI agents study Parkinson-like disease data in 32 minutes","summary":"OpenAI, in collaboration with PerturbAI and Scripps Research, announced a new study demonstrating the use of thousands of AI agents to analyze experimental data from millions of mouse brain cells. The team utilized GPT-Rosalind, OpenAI’s specialized life sciences model, to investigate how specific gene disruptions affect cellular responses. Each agent focused on a single genetic change, searching scientific literature for explanations and compiling findings into a \"reasoning atlas.\" The entire analysis, which could take months to a year manually, was completed in 32 minutes.\n\nThe study identified potential links between cholesterol regulation and neurodegeneration, specifically noting increased cholesterol-producing gene activity when the DCTN1 gene (associated with Perry syndrome) was disrupted. The agents also detected unexpected differences in how Tsc1 and Tsc2 genes affect cerebellum neurons. While these are early research leads requiring further testing, they offer new clues for neurological diseases like Parkinson's.\n\nThe researchers compared the distributed agent approach against a single AI system. The single agent cited only 11 scientific references, whereas the distributed agents cited over 700. Human experts, unaware of the method, preferred the distributed-agent reports 75% of the time. This suggests that deploying many agents for smaller tasks can uncover connections in large datasets that a single model might miss.","keyPoints":["OpenAI agents analyzed mouse brain cell data in 32 minutes, a task taking months manually.","Distributed agents cited over 700 references, while a single agent cited only 11.","Human experts preferred the distributed-agent reports 75% of the time in blind tests."],"whyItMatters":"This demonstrates how distributed AI agents can accelerate medical research by rapidly connecting scattered scientific data, potentially identifying new disease mechanisms faster than manual review or single-model analysis.","category":{"slug":"research","name":"Research","url":"https://digestai.news/category/research"},"entities":{"companies":["OpenAI","PerturbAI","Scripps Research","University of California, San Francisco"],"models":["GPT-Rosalind"],"people":["Tomasz Nowakowski"]},"firstPublishedAt":"2026-10-21T23:45:00Z","updatedAt":"2026-10-21T23:45:00Z","sourceCount":1,"hasPrimarySource":false,"sources":[{"outlet":"tech.yahoo.com","title":"OpenAI unleashed thousands of AI agents to study a Parkinson-like disease — here's what they found in 32 minutes","url":"https://tech.yahoo.com/ai/articles/openai-unleashed-thousands-ai-agents-064500385.html","publishedAt":"2026-10-21T23:45:00Z","type":"press","primary":false,"lead":true}],"sourceNotes":null,"discussions":[],"thread":{"title":"AI Agents Accelerate Brain Research","url":"https://digestai.news/thread/openai-perturbai-and-scripps-research-use-gptrosalind-agents-to-analyze-mouse","storyCount":2},"cite":{"text":"Digest AI, \"OpenAI agents study Parkinson-like disease data in 32 minutes\", 21 October 2026, https://digestai.news/story/openai-agents-study-parkinson-like-disease-data-in-32-minutes","publisher":"Digest AI","title":"OpenAI agents study Parkinson-like disease data in 32 minutes","datePublished":"2026-10-21T23:45:00Z","url":"https://digestai.news/story/openai-agents-study-parkinson-like-disease-data-in-32-minutes"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}