Abstract:

The application of computers in enzyme engineering has led to the continuous expansion of the sequence space exploration of enzymes. With the establishment of different molecular force fields, many algorithms came out upon computing molecular energy and were applied to the enzyme redesign and screening of catalytic activity, stability, substrate specificity, etc. With the improvement of computer hardware and the optimization of algorithms, artificial enzymes with completely new functions have been successfully designed and developed. In recent years, artificial intelligence has made breakthroughs in protein structure prediction and has also been applied to enzyme design. Basis of molecular force field and enzyme design and selection algorithm were introduced in this paper. The methods and successful cases of de novo design, as well as the process of machine learning to design enzymes and the latest research progress are described emphatically. The outlooks of artificial intelligence in the enzyme engineering are given in the end, contributing for enzymatic engineering and brand-new functional biocatalysts design.

Key words: enzyme engineering, de novo design, artificial intelligence, mechanism, energy function

摘要：

计算机在酶工程中的应用使得酶的序列空间探索度不断被扩大。随着不同分子力场参数的建立，涌现出诸多以计算分子能量为基础的算法，并被用于酶的催化活性、稳定性、底物特异性等的改造与筛选。伴随计算机硬件的提升与算法的优化，从头设计全新功能的人工酶取得成功并得以发展。近年来，人工智能在蛋白质结构预测上不断获得突破，同时也被应用到酶的设计中。介绍了分子力场基础和酶设计与筛选的算法，重点阐述了从头设计的方法和成功案例，以及机器学习设计酶的流程和最新的研究进展，展望了人工智能在酶工程领域的未来发展，为酶的改造与全新功能的生物催化剂的设计助力。

关键词: 酶工程, 从头设计, 人工智能, 机理, 能量函数