文献类型：期刊文献

英文题名：Deep learning approaches for de novo drug design: An overview

作者：Wang, Mingyang[1];Wang, Zhe[1];Sun, Huiyong[2];Wang, Jike[1];Shen, Chao[1];Weng, Gaoqi[1];Chai, Xin[1];Li, Honglin[1,3];Cao, Dongsheng[4];Hou, Tingjun[1]

机构：[1]Zhejiang Univ, Coll Pharmaceut Sci, Innovat Inst Artificial Intelligence Med, Hangzhou 310058, Zhejiang, Peoples R China;[2]China Pharmaceut Univ, Dept Med Chem, Nanjing 210009, Jiangsu, Peoples R China;[3]East China Univ Sci & Technol, Sch Pharm, Shanghai Key Lab New Drug Design, Shanghai 200237, Peoples R China;[4]Cent South Univ, Xiangya Sch Pharmaceut Sci, Changsha 410013, Hunan, Peoples R China

年份：2022

卷号：72

起止页码：135

外文期刊名：CURRENT OPINION IN STRUCTURAL BIOLOGY

收录：;WOS:【SCI-EXPANDED(收录号:WOS:000768730600016)】；

基金：This work was financially supported by the National Natural Science Foundation of China (21575128, 81773632) , Natural Science Foundation of Zhejiang Province (LZ19H300001) , Key R&D Program of Zhejiang Prov-ince (2020C03010) , and Fundamental Research Funds for the Central Universities (2020QNA7003) .

语种：英文

摘要：De novo drug design is the process of generating novel lead compounds with desirable pharmacological and physiochemical properties. The application of deep learning (DL) in de novo drug design has become a hot topic, and many DLbased approaches have been developed for molecular generation tasks. Generally, these approaches were developed as per four frameworks: recurrent neural networks; encoderdecoder; reinforcement learning; and generative adversarial networks. In this review, we first introduced the molecular representation and assessment metrics used in DL-based de novo drug design. Then, we summarized the features of each architecture. Finally, the potential challenges and future directions of DL-based molecular generation were prospected.