Premium
Can Machine Learning Revolutionize Directed Evolution of Selective Enzymes?
Author(s) -
Li Guangyue,
Dong Yijie,
Reetz Manfred T.
Publication year - 2019
Publication title -
advanced synthesis and catalysis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.541
H-Index - 155
eISSN - 1615-4169
pISSN - 1615-4150
DOI - 10.1002/adsc.201900149
Subject(s) - directed evolution , artificial intelligence , machine learning , protein engineering , directed molecular evolution , chemistry , saturated mutagenesis , training set , mutagenesis , computer science , enzyme , mutation , biochemistry , mutant , gene
Machine learning as a form of artificial intelligence consists of algorithms and statistical models for improving computer performance for different tasks. Training data are utilized for making decisions and predictions. Since directed evolution of enzymes produces huge amounts of potential training data, machine learning seems to be ideally suited to support this protein engineering technique. Machine learning has been used in protein science for a long time with different purposes. This mini‐review focuses on the utility of machine learning as an aid in the directed evolution of selective enzymes. Recent studies have shown that the algorithms ASRA and Innov'SAR are well suited as guides when performing saturation mutagenesis at sites lining the binding pocket for enhancing stereoselectivity and activity.