Recently, Yan Ning, a famous structural biologist, announced that he would resign from Princeton University and set up the Academy of Medical Sciences in Shenzhen. In response to the latest research that she will focus on when she returns to China, she said that the discovery of structural biology is of great significance to the pharmaceutical industry, including the research on the interaction between drugs and hormones, which has more implications for drug development and disease treatment.
Yan Ning said, for example, that through the method of structural biology, scientists saw for the first time that two commonly used drugs used protein as a "scaffold", which directly "side by side" affected the normal function of protein.
Technological change promotes the development of structural biology.
Protein, as the scaffold and main substance of human tissues and organs, plays an important role in human life activities. In the cell, a large number of protein elements form the molecular machines, which performs important molecular processes in the cell through the interaction of protein, including the cell’s response to the external environment and the internal environment, and also forms a signal transduction network system with the interaction of protein as a link.
Structure By studying the three-dimensional structure, dynamic process and biological function of biological macromolecules, biology can provide more details of protein interaction and real-time dynamic process, thus helping scientists to understand molecular mechanism and explore the pathogenesis of diseases related to macromolecular dysfunction.
The development of biology not only contributes to the discovery of drugs, but also affects the research fields of life sciences including biochemistry, cell biology, genetic development, neurobiology, microbiology and pathopharmacology.
The development of science and technology has been promoting the progress of structural biology. In 2017, the Nobel Prize in Chemistry was awarded to cryoelectron microscopy, which can greatly improve the efficiency of analyzing the atomic resolution three-dimensional structure of large protein complexes; Moreover, researchers can freeze biomolecules in motion and visualize their motion process.
This breakthrough has brought a perfect storm to the field of structural biology. In recent years, important breakthroughs have been made in the field of life sciences. The work of China scientists such as Cheng Yifan, Shi Yigong, Yang Maojun and Liu Zhengfeng has also benefited from this technology. They have analyzed the important complex structure of atomic resolution. In addition, the enzyme that produces the protein that leads to Alzheimer’s syndrome was analyzed by cryomicroscope.
In August, 2015, Shi Yigong’s research team published an article in Nature, reporting the three-dimensional electron microscope structure of human γ -secretase with a resolution of 3.4 angstrom, and studying the function of the pathogenic mutant of γ -secretase based on the structure analysis, which provided an important foundation for understanding the working mechanism of γ -secretase and the pathogenesis of Alzheimer’s disease.
In February, 2020, after the outbreak of COVID-19, the research team of West Lake University successfully analyzed the full-length structure of novel coronavirus receptor ACE2 for the first time by using cryo-electron microscopy, which helped the research and development of drugs in COVID-19.
What does AI predict protein folding change?
With the help of artificial intelligence technology, DeepMind, a subsidiary of Google, recently announced 220 million kinds of protein structures predicted by AlphaFold software, which shocked the field of structural biology, because it indicates that artificial intelligence enterprises have begun to "truly hand over the power of Al to scientists all over the world".
Scientists compare the significance of this disruptive breakthrough with the human genome project. In the 1990s, when the Human Genome Project began to take shape, scientists realized that it was not enough to master the base arrangement of genes, but also to know the protein, the product of genes.
Shi Yigong, academician of China Academy of Sciences, structural biologist and president of West Lake University, commented on AlphaFold’s work: "Its accurate prediction of protein structure is the greatest contribution of artificial intelligence to the scientific field, and it is also one of the most important scientific breakthroughs made by mankind in the 21st century."
Shi Yigong once told China Business News: "AlphaFold represents the world’s leading artificial intelligence protein organization prediction system." At the same time, he said, China’s high-tech enterprises are catching up, expecting to bring surprises to the world in the near future.
He also said that the improvement of the accuracy of protein structure prediction will greatly benefit the pharmaceutical industry. "Artificial intelligence prediction of protein structure provides an important foundation for drug design and optimization. The structure of drug target proteins combined with all small molecule drugs can almost be wiped out by AlphaFold. " Shi Yigong said.
In the opinion of some scientists, although AlphaFold’s work is shocking, the accuracy of drug research and development prediction is not enough. Professor Liu Zhijie, executive director of iHuman Institute, Shanghai University of Science and Technology, told the First Financial Reporter: "It has been a long time to predict the protein structure. Now the accuracy of the prediction is definitely getting higher and higher, but it still hasn’t reached the precision of crystal structure."
Liu Zhijie First Financial News told the reporter that the crystal structure is the most accurate, and now artificial intelligence can predict the protein folding with the accuracy of electron microscope and nuclear magnetic resonance. In addition, because there are thousands of structures in protein, the difficulty of analysis is different. "If some protein sequences are similar to the known structure of artificial intelligence, it is easier to predict." Liu Zhijie said.
However, he still believes that with the continuous improvement of the prediction accuracy of protein folding, it will play a more important role in the field of life sciences in the future. "If the current prediction can reach the accuracy of electron microscope, some drugs can already be designed. Drug design is the biggest application field of artificial intelligence protein folding prediction." Said to Liu Zhijie First Financial Reporter.
Others think that with the development of artificial intelligence, there may not be so many structural biologists in the future. "Many researchers who do structural biology are actually more like technical service personnel. The more people there are, the more structures that can be analyzed. Therefore, in essence, a large part of their work depends on manpower. Now with AI, it is true that a large number of people who do structural biology have changed careers." A virus researcher told the First Financial News reporter.
However, for top structural biologists such as Shi Yigong, technology is only a powerful support for the "top brain", which can help them realize more ideas. "Every conscientious biologist should know how to make good use of the structural prediction of artificial intelligence." Shi Yigong said.
Professor Michael Levitt, winner of the Nobel Prize in Chemistry in 2013, told China Business News: "I think many structural biologists are not only doing structural research, but also doing a lot of work on protein function and drug research and development, just like Professor Yan Ning. Artificial intelligence only liberates a part of traditional manpower, but the progress of science still needs to rely on the smartest human brain. I’m afraid that artificial intelligence alone will not work. "
In recent years, a number of AI pharmaceutical companies have also been born in China. In this regard, Ewing Nan, academician of Chinese Academy of Sciences and president of Beijing Institute of Scientific Intelligence, pointed out to the First Financial Reporter: "The emergence of AI for Science research paradigm is an important historical opportunity for scientific and technological innovation, which not only expands the capability boundary of data-driven and physical model-driven models, but also is expected to promote the organic combination of the two, provide theoretical basis for further solving practical problems, and greatly narrow the distance between scientific research and practical application."
