Deep learning methods for molecular representation and property prediction

被引：140

作者：

Li, Zhen ^{[1
]}

Jiang, Mingjian ^{[3
]}

Wang, Shuang ^{[4
]}

Zhang, Shugang ^{[2
]}

机构：

[1] Qingdao Univ, Coll Comp Sci & Technol, Qingdao 266071, Shandong, Peoples R China

[2] Ocean Univ China, Coll Comp Sci & Technol, Qingdao 266100, Shandong, Peoples R China

[3] Qingdao Univ Technol, Sch Informat & Control Engn, Qingdao 266033, Shandong, Peoples R China

[4] China Univ Petr, Coll Comp Sci & Technol, Qingdao 266580, Shandong, Peoples R China

来源：

DRUG DISCOVERY TODAY | 2022年 / 27卷 / 12期

关键词：

Molecular representation; Deep learning; Self-supervised learning; Drug discovery; Property prediction; PERFORMANCE; NETWORK; SMILES; INFORMATION; TOPOLOGY;

D O I：

10.1016/j.drudis.2022.103373

中图分类号：

R9 [药学];

学科分类号：

1007 ;

摘要：

With advances in artificial intelligence (AI) methods, computer-aided drug design (CADD) has developed rapidly in recent years. Effective molecular representation and accurate property prediction are crucial tasks in CADD workflows. In this review, we summarize contemporary applications of deep learning (DL) methods for molecular representation and property prediction. We categorize DL methods according to the format of molecular data (1D, 2D, and 3D). In addition, we discuss some common DL models, such as ensemble learning and transfer learning, and analyze the interpretability methods for these models. We also highlight the challenges and opportunities of DL methods for molecular representation and property prediction.

引用

页数：13

共 117 条

[1]

[Anonymous], RDKit: Open-Source Cheminformatics Software

[2] Calibrated uncertainty for molecular property prediction using ensembles of message passing neural networks [J].

Busk, Jonas ;

Bjorn Jorgensen, Peter ;

Bhowmik, Arghya ;

Schmidt, Mikkel N. ;

Winther, Ole ;

Vegge, Tejs .

MACHINE LEARNING-SCIENCE AND TECHNOLOGY, 2022, 3 (01)

[3] Integration of element specific persistent homology and machine learning for protein-ligand binding affinity prediction [J].

Cang, Zixuan ;

Wei, Guo-Wei .

INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN BIOMEDICAL ENGINEERING, 2018, 34 (02)

[4] Representability of algebraic topology for biomolecules in machine learning based scoring and virtual screening [J].

Cang, Zixuan ;

Mu, Lin ;

Wei, Guo-Wei .

PLOS COMPUTATIONAL BIOLOGY, 2018, 14 (01)

[5] TopologyNet: Topology based deep convolutional and multi-task neural networks for biomolecular property predictions [J].

Cang, Zixuan ;

Wei, Guowei .

PLOS COMPUTATIONAL BIOLOGY, 2017, 13 (07)

[6] Prediction of Energetic Material Properties from Electronic Structure Using 3D Convolutional Neural Networks [J].

Casey, Alex D. ;

Son, Steven F. ;

Bilionis, Ilias ;

Barnes, Brian C. .

JOURNAL OF CHEMICAL INFORMATION AND MODELING, 2020, 60 (10) :4457-4473

[7] Different molecular enumeration influences in deep learning: an example using aqueous solubility [J].

Chen, Jen-Hao ;

Tseng, Yufeng Jane .

BRIEFINGS IN BIOINFORMATICS, 2021, 22 (03)

[8]

Chen T, 2020, PR MACH LEARN RES, V119

[9] MRlogP: Transfer Learning Enables Accurate logP Prediction Using Small Experimental Training Datasets [J].

Chen, Yan-Kai ;

Shave, Steven ;

Auer, Manfred .

PROCESSES, 2021, 9 (11)

[10] Enhanced Deep-Learning Prediction of Molecular Properties via Augmentation of Bond Topology [J].

Cho, Hyeoncheol ;

Choi, Insung S. .

CHEMMEDCHEM, 2019, 14 (17) :1604-1609

← 1 2 3 4 5 6 7 8 9 10 →