文献类型：期刊文献

英文题名：SESNet: sequence-structure feature-integrated deep learning method for data-efficient protein engineering

作者：Li, Mingchen[1,2,6];Kang, Liqi[1,2,4,5];Xiong, Yi[7];Wang, Yu Guang[1,2,3];Fan, Guisheng[6];Tan, Pan[1,2,3];Hong, Liang[1,2,3,4,5]

机构：[1]Shanghai Jiao Tong Univ, Shanghai Natl Ctr Appl Math, Shanghai 200240, Peoples R China;[2]Shanghai Jiao Tong Univ, Inst Nat Sci, Shanghai 200240, Peoples R China;[3]Shanghai Artificial Intelligence Lab, Shanghai 200240, Peoples R China;[4]Shanghai Jiao Tong Univ, Sch Phys & Astron, Shanghai 200240, Peoples R China;[5]Shanghai Jiao Tong Univ, Sch Pharm, Shanghai 200240, Peoples R China;[6]East China Univ Sci & Technol, Sch Informat Sci & Engn, Shanghai 200240, Peoples R China;[7]Shanghai Jiao Tong Univ, Sch Life Sci & Biotechnol, Shanghai 200240, Peoples R China

年份：2023

卷号：15

期号：1

外文期刊名：JOURNAL OF CHEMINFORMATICS

收录：;WOS:【SCI-EXPANDED(收录号:WOS:000926216400002)】；

基金：This work was financially supported by the Natural Science Foundation of China (Grant No. 12104295, 11974239, 31630002, 61872094, 61832019), the Innovation Program of Shanghai Municipal Education Commission, and Shanghai JiaoTong university Multidisciplinary research fund of medicine and engineering YG 2016QN13.The main work was supported by Shanghai Artificial Intelligence Laboratory, and the computing hardware resource was also mainly supported by them. While some testing were supported by the Center for High Performance Computing at Shanghai JiaoTong University.

语种：英文

摘要：Deep learning has been widely used for protein engineering. However, it is limited by the lack of sufficient experimental data to train an accurate model for predicting the functional fitness of high-order mutants. Here, we develop SESNet, a supervised deep-learning model to predict the fitness for protein mutants by leveraging both sequence and structure information, and exploiting attention mechanism. Our model integrates local evolutionary context from homologous sequences, the global evolutionary context encoding rich semantic from the universal protein sequence space and the structure information accounting for the microenvironment around each residue in a protein. We show that SESNet outperforms state-of-the-art models for predicting the sequence-function relationship on 26 deep mutational scanning datasets. More importantly, we propose a data augmentation strategy by leveraging the data from unsupervised models to pre-train our model. After that, our model can achieve strikingly high accuracy in prediction of the fitness of protein mutants, especially for the higher order variants (> 4 mutation sites), when finetuned by using only a small number of experimental mutation data (< 50). The strategy proposed is of great practical value as the required experimental effort, i.e., producing a few tens of experimental mutation data on a given protein, is generally affordable by an ordinary biochemical group and can be applied on almost any protein.