{"version":1,"type":"story","url":"https://digestai.news/story/abhinav-adduri-and-colleagues-develop-state-model-for-generalizable","json":"https://digestai.news/story/abhinav-adduri-and-colleagues-develop-state-model-for-generalizable.json","markdown":"https://digestai.news/story/abhinav-adduri-and-colleagues-develop-state-model-for-generalizable.md","slug":"abhinav-adduri-and-colleagues-develop-state-model-for-generalizable","headline":"Abhinav Adduri and colleagues develop State model for generalizable perturbation prediction","summary":"Researchers Abhinav Adduri, Yusuf Roohani and their team introduced State, a machine‑learning system designed to predict how cells respond to perturbations across different biological contexts. State combines a transformer‑based state transition (ST) module that forecasts single‑cell perturbation effects with a state embedding (SE) module pretrained on a wide range of single‑cell profiles.\n\nThe ST model was trained on perturbation data from over 100 million cells, while the SE model learned from observational data covering 167 million cells, giving the system broad knowledge of cellular states. To benchmark State, the authors also built Cell‑Eval, an evaluation framework that goes beyond standard accuracy metrics to assess perturbation distinction, differential‑gene identification, expression‑change prediction and overall perturbation estimation. The work appears in Nature Machine Learning as subscription‑only content.","keyPoints":["Adduri, Roohani et al. released State, a transformer‑based model for single‑cell perturbation prediction.","State was trained on >100 million perturbation cells and >167 million observational cells.","Cell‑Eval framework evaluates accuracy, perturbation distinction, DE‑gene detection and expression change prediction."],"whyItMatters":"Predicting perturbation effects across contexts can speed drug target validation and functional genomics, reducing costly experimental work.","category":{"slug":"models","name":"Generative AI & Models","url":"https://digestai.news/category/models"},"entities":{"companies":[],"models":["State"],"people":["Abhinav Adduri","Yusuf Roohani"]},"firstPublishedAt":"2026-10-07T00:00:00Z","updatedAt":"2026-10-07T00:00:00Z","sourceCount":1,"hasPrimarySource":true,"sources":[{"outlet":"Nature Machine Learning","title":"Generalizable perturbation prediction","url":"https://nature.com/articles/s41592-026-03260-1","publishedAt":"2026-10-07T00:00:00Z","type":"primary","primary":true,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"Abhinav Adduri and colleagues develop State model for generalizable perturbation prediction\", 7 October 2026, https://digestai.news/story/abhinav-adduri-and-colleagues-develop-state-model-for-generalizable","publisher":"Digest AI","title":"Abhinav Adduri and colleagues develop State model for generalizable perturbation prediction","datePublished":"2026-10-07T00:00:00Z","url":"https://digestai.news/story/abhinav-adduri-and-colleagues-develop-state-model-for-generalizable"},"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"}