echemi logo
  • Product
  • Supplier
  • Inquiry
    Home > Biochemistry News > Biotechnology News > The team of Zhao Yilei of Shanghai Jiaotong University proved the effectiveness of biosynthase pre-reactive state analysis based on artificial intelligence

    The team of Zhao Yilei of Shanghai Jiaotong University proved the effectiveness of biosynthase pre-reactive state analysis based on artificial intelligence

    • Last Update: 2022-11-14
    • Source: Internet
    • Author: User
    Search more information of high quality chemicals, good prices and reliable suppliers, visit

    Recently, Cell Press Cell Reports Physical Science published online the research team of Professor Zhao Yilei of the School of Life Science and Technology of Shanghai Jiao Tong University entitled "Understanding the effectiveness of enzyme pre-reaction state by quantum-based machine learning.
    " model"
    Luo Shenggan, a doctoral student in the School of Life Science and Technology, is the first author of the paper, and Liu Lanxuan of Shanghai Artificial Intelligence Laboratory assisted in the machine learning modeling work
    In this study, the stereoselective prediction of chiral alcohol products in yeast alcohol dehydrogenase CpRCR protein engineering is taken as the application scenario, and the relative stability of the two pre-reactive conformations under the Proleg rule and the anti-Proleg rule is measured by the traditional umbrella sampling method, and it is found that the simulation calculation results are basically consistent
    with the experimental observations 。 The authors further use the high-order quantum chemistry QM/MM method to collect more than 700 pairs of three-dimensional structural information of "pre-reactive state-transition state", use more than 1,000 topological feature parameters of catalytic reaction centers to carry out artificial intelligence machine learning LASSO-SVM regression analysis, and finally explain why the pre-reactive state QML model with only dozens of selected parameters can effectively predict the catalytic reaction activity
    of CpRCR enzymes based on the front-line reaction orbital theory.

    Enzyme-catalyzed stereoselectivity prediction is a major challenge in scientific engineering computing, which requires high-order quantum mechanical methods to measure small differences
    between different reaction pathways in highly complex macromolecular systems.
    The rapid molecular dynamics simulation method based on the pre-reactive state has been widely used in the analysis of enzyme catalytic activity, but there have been various controversies for a long time on how to select the pre-reactive state topology parameters of the near-attack conformation NAC, and it is urgent to break through the quantitative effect
    of transition state and pre-reactive state on enzyme catalytic activity in theory 。 Based on the accumulation of scientific computing of enzyme engineering in the long-term cooperation with Nie Yao's research group of the Bioengineering Laboratory of Jiangnan University, the research group selected the alcohol dehydrogenase CpRCR with clear and reliable reaction mechanism, rich residue mutation experimental data, and wide application in bioindustrial production as the research entry point, and used the pro-R and pro-S active conformations obtained in the molecular dynamics pre-reactive state simulation of umbrella sampling to carry out large-scale quantum mechanical calculations in the High Performance Computing Center of Shanghai Jiao Tong University.
    MM structure data-driven method constructs a high-precision artificial intelligence machine learning model
    with 99.
    6% explanatory power based on "pre-reactive-transition state" joint data.
    Then, two QML comparison models based on independent datasets of "pre-reactive state" or "transition state" were constructed using the same modeling process, and it was found that the enzyme activity prediction ability of PRS-QML was as high as 90.
    7%, while the enzyme activity prediction ability of TS-QML was only 55.
    4%, which successfully demonstrated the overwhelming advantage
    of pre-reactive state model prediction in practical application scenarios such as biosynthetic enzyme residue mutation and substrate modification 。 The authors further use the frontline molecular orbital theory to illustrate the correlation between the structural topological parameters enriched in the QML model and the primary-secondary orbital interaction, and clarify that the influence of protein engineering on the potential energy surface of enzyme catalytic reactions is mainly concentrated in the changes of the pre-reactive state region, while the transition state region of biosynthase in natural or artificial evolution has been highly optimized and relatively stable

    This research work is another breakthrough achievement of Zhao Yilei's research group in the long-term research of reaction pathway calculation system of complex systems, which will significantly promote the application
    of pre-reactive state models in the field of biosynthase protein engineering transformation.

    The research was supported
    by the National Key R&D Program "Synthetic Biology" and the National Natural Science Foundation of China.

    College of Life Science and Technology
    College of Life Science and Technology
    This article is an English version of an article which is originally in the Chinese language on and is provided for information purposes only. This website makes no representation or warranty of any kind, either expressed or implied, as to the accuracy, completeness ownership or reliability of the article or any translations thereof. If you have any concerns or complaints relating to the article, please send an email, providing a detailed description of the concern or complaint, to A staff member will contact you within 5 working days. Once verified, infringing content will be removed immediately.

    Contact Us

    The source of this page with content of products and services is from Internet, which doesn't represent ECHEMI's opinion. If you have any queries, please write to It will be replied within 5 days.

    Moreover, if you find any instances of plagiarism from the page, please send email to with relevant evidence.