Heavy weight! DeepMind is scaling up again!

AlphaFold dominates the field of protein prediction, and AlphaFold 3 breaks through its limitations by predicting all living molecules. When it was released, there were voices claiming that AlphaFold 3's structured prediction and generation would accelerate the development of AI driven drug design.

Now, DeepMind has scaled up its strategy by releasing AlphaProteo for the design of new proteins.

AlphaProteo is the first AI system designed specifically for designing novel high-intensity protein conjugates, capable of generating conjugates of multiple target proteins. This system significantly improves the success rate, even 3 to 300 times stronger than existing methods. Especially for the VEGF-A protein related to cancer and diabetes complications, AlphaProteo is the first AI tool to design successful conjugates. In addition, it can effectively bind to various proteins related to infection, cancer, inflammation, and autoimmune diseases, such as BHRF1, SARS-CoV-2 spike protein, IL-7R α, PD-L1, TrkA, and IL-17A.

AlphaProteo has successfully designed protein binding agents for 7 targets. Including two viral proteins, BHRF1 and SARS-CoV-2 spike protein receptor binding domain SC2RBD; And five proteins involved in cancer, inflammation, and autoimmune diseases, SC2RBD, IL-7R α, PD-L1, TrkA, IL-17A, and VEGF-A. Achieved higher experimental success rates and binding affinities 3 to 300 times higher than existing best methods.

For BHRF1, 88% of the candidate molecules successfully bound when tested in Google DeepMind Wet Lab. According to the tested targets, the binding affinity of AlphaProteo binding agent is on average 10 times stronger than the existing optimal design method; For another target TrkA, the binding agent designed by the team is even stronger than the best previously designed binding agent optimized for this target after multiple rounds of experiments!

As the most anticipated technology of 2024! There are very few materials and learning paths for AI protein design, and special training and learning are urgently needed! Zhengzhou Qingrui Information Technology Co., Ltd., in collaboration with Tsinghua University, Peking University, Xihu University, Zhejiang University, University of Science and Technology of China, Tianjin University, and Peking Union Medical College Pharmaceutical Research Institute, has held over 40 training sessions with more than 5000 attendees! Highly praised by students! Among them, there are publications in top international journals such as Nature, Cell, and Science!