OpenAI says GPT-Rosalind offers new leads for neurodegenerative disease drug discovery

The Decoder

OpenAI agents complete mouse brain data analysis in 32 minutes, a task that takes humans months

OpenAI, in collaboration with PerturbAI and Scripps Research, used thousands of AI agents based on GPT-Rosalind to analyze a whole-brain CRISPR atlas covering millions of mouse cells. The entire analysis took 32 minutes, and the team estimates that doing it manually would take months to a year. Each agent was responsible for one gene perturbation, analyzing cellular responses and searching the literature for explanations. The results revealed an unexpected difference between two disease-related genes, as well as a possible link between cholesterol regulation and neurodegeneration. In blind tests, human experts preferred the reports from distributed agents over those from a single agent in 75% of cases; the generated Reasoning Atlas has been made public on GitHub, and the accompanying paper was published on bioRxiv.

Unite.AISelected

OpenAI, PerturbAI, and Scripps use GPT-Rosalind agents to build a mouse brain reasoning atlas

On October 21, 2026, OpenAI announced that, with PerturbAI and Scripps Research, it used thousands of AI agents driven by GPT-Rosalind to analyze existing experimental data from millions of mouse brain cells within 32 minutes and aggregated the reports of each agent into a public reasoning atlas. The method is called distributed biological reasoning, with each agent responsible for one gene perturbation, analyzing cellular responses, retrieving literature, and proposing explanations. The team says equivalent manual analysis and literature review would take months to a year; under fixed models and resources, a single agent handling the maximum workload cited only 11 papers, while distributed agents cited more than 700, and in blind evaluation human experts preferred the distributed agent reports in 75% of cases. A companion preprint was released on bioRxiv on October 16, 2026, and has not yet undergone peer review; the reasoning atlas and the Distributed Agents framework are publicly available on GitHub under the MIT license, and the dataset is the whole-brain CRISPR Perturb-seq atlas from PerturbAI on Hugging Face.

Rohan PaulSelected

OpenAI and others use 1,946 agents to interpret a mouse whole-brain CRISPR atlas in 32 minutes

OpenAI, PerturbAI, and Scripps Research used 1,946 GPT-Rosalind-5.5 agents to complete interpretation of a mouse whole-brain CRISPR atlas in 32 minutes; OpenAI estimates that manual work would take months. Each agent analyzed only the effect of a single gene knockout and retrieved literature, and did not access the atlas paper published after the model's December 2025 knowledge cutoff, yet still matched or refined 48 of 55 published findings supportable by the expression data. The core of the method is to split one giant analysis into nearly 2,000 single-gene small tasks executed in parallel; a single agent processing 893 genes cited only 11 papers, while after splitting the citation count was 711 times that, and blinded PhD reviewers preferred the split reports in 75% of cases.

openaiFirst-party

English translation is being prepared.

Phoenix Technology

PerturbAI and OpenAI Jointly Develop a Parallel Research Framework with Nearly 2,000 AI Agents

Biotechnology company PerturbAI announced a collaboration with OpenAI and Scripps Research to develop an AI reasoning framework in which nearly 2,000 AI agents work in parallel to analyze complex biological experimental data, search the literature, and propose testable hypotheses. When analyzing whole-brain CRISPR maps, the system completed the integration of data and research findings in 32 minutes; the company estimates that humans working serially would need months to a year. In blind-review comparisons against reports from a single AI agent, experts preferred the system's explanations 75% of the time. The collaboration has lasted about a year, and next the two parties will jointly conduct new drug target discovery using causal biological data from living systems; the relevant leads still require follow-up experimental validation.