science π¬ AI models could speed up drug discovery using digital cells
Researchers developed two digital models using 4D lattice light-sheet microscopy to predict mitochondrial responses to drugs. These models were created at the University of California San Diego, CA, USA. The first AI, MitoSpace, was trained on 40,000 4D cancer cell videos with 75% accuracy. The second model created a 'digital twin' of a human carcinoma cell to mimic real mitochondrial transport. The team plans to combine these approaches to eventually model whole tissues for better clinical insights. π¬