浙大发布高精准基因组设计AI模型

Core Insights - The "Nüwa CE" deep learning AI model developed by Professor Guo Guoji's team at Jiang University can predict phenotypic changes due to mutations in genomic regulatory regions with over 90% accuracy [1][2] - The model is based on a high-throughput, high-sensitivity single-nucleus chromatin accessibility sequencing technology, creating a comprehensive regulatory element atlas covering five representative vertebrate species [1] - The model aims to enhance understanding of the complex regulatory language hidden within vast gene sequences, contributing to advancements in life sciences, medicine, and agriculture [1][2] Group 1 - The "Nüwa CE" model surpasses existing genomic AI models in multiple metrics, demonstrating its capability to predict phenotypic changes in various cell types following mutations in genomic regulatory elements [2] - The model has been applied to modify a therapeutic gene locus for sickle cell anemia, resulting in increased fetal hemoglobin expression [2] - The development of the model is part of a broader effort to decode the genetic information that remains largely unexplored, with less than 10% of the human genome's hereditary information deciphered [1]