Scientists have discovered that using artificial intelligence can efficiently predict protein molecular structure

Recently, scientists from Google in the United States developed and utilized the AlphaFold system to efficiently predict the molecular structures of 350000 proteins through artificial intelligence. The relevant research results have been published in Nature, titled: Highly accurate protein structure prediction for the human project.

Analyzing the molecular structure of proteins is of great significance for basic research, drug development, and disease diagnosis and treatment. Currently, scientists have only determined the molecular structure of 17% of human proteins through traditional techniques such as X-ray crystallography, nuclear magnetic resonance, or cryo electron microscopy. This study made molecular structure predictions for a total of 350000 proteins from 20 commonly used model organisms in biological research, including humans. Among them, 20296 are human proteins, accounting for approximately 98.5% of the total human proteins.

This study innovatively utilizes artificial intelligence technology to achieve disruptive innovation in protein molecular structure analysis, greatly improving the efficiency and scope of research, bringing new research tools to life science research, and opening up a new mode of research in the fields of life science and biomedicine.

Prev: Scientists use AI technology to obtain clearer images of lunar craters