According to a report on the official website of the University of Basel in Switzerland on the 20th, scientists from the university and the Swiss Institute of Bioinformatics (SIB) have identified 290 new protein families and a new protein fold resembling a flower shape using machine learning technology. The relevant paper was published in the recent journal Nature. In the past few years, 'alpha folding' has completely changed protein science. This artificial intelligence tool utilizes protein data collected by scientists over the past 50 years for training, and can accurately predict the 3D shape of proteins. Last year, 215 million proteins were modeled, providing information on almost any protein shape, helping scientists better understand the evolution process of proteins. However, in the field of protein research, there are still many secrets waiting to be revealed. The research team has constructed an interactive network consisting of 53 million proteins, the Protein Universe Atlas, to better predict unknown protein families and their functions on a large scale. On this basis, the research team identified 290 new protein families and a new protein folding resembling a flower shape. Researchers point out that using deep learning based tools to search for unknown proteins in the "protein universe map" paves the way for innovation in life sciences. The "Protein Universe Map" helps to study more proteins, and understanding the structure and function of proteins is the first step in developing new drugs or altering protein function through gene editing tools. (New News Agency)