基于深度学习的药物设计方法

doi:10.3969/j.issn.0253-9608.2021.05.009

摘要/Abstract

摘要： 药物研发是一个长周期、高投入和高风险的过程。随着科学技术的不断革新，生物医学数据呈爆炸式增长，为深度学习技术在生物医药领域的应用带来了契机，同时也为加速新药研发赋予了前所未有的希望。文章围绕药物设计流程，简要介绍深度学习算法在药物靶标发现、分子生成、基于配体的药物设计和基于结构的药物设计4个主要环节中的应用和研究进展。

关键词: 药物设计, 人工智能, 深度学习, 大数据

Abstract: Drug design and development is a long term, expensive and high-risk process. The continuous innovation of science and technology, as well as the explosive growth of the biomedical data brings unprecedented opportunities for the applications of deep learning algorithms in biomedical areas to expediate the drug development. In this review, we briefly overview and list the state-ofthe-art deep learning-driven drug design methods for identifying drug targets, molecular generation, ligand-based and structure-based drug design.

Key words: , drug design, artificial intelligence, deep learning, big data

李风雷, 胡乔宇, 熊若凡, 白芳. 基于深度学习的药物设计方法[J]. 自然杂志, 2021, 43(5): 383-390.

LI Fenglei, HU Qiaoyu, XIONG Ruofan, BAI Fang. Computational drug design methods by deep learning algorithms [J]. Chinese Journal of Nature, 2021, 43(5): 383-390.

参考文献

[1] VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need [J]. NIPS, 2017, 17: 6000-6010.

[2] DEVLIN J, CHANG M-W, LEE K, et al. BERT: pre-training of deep bidirectional transformers for language understanding [J]. NAACL-HLT, 2019, 1: 4171-4186.

[3] RAMSUNDAR B, EASTMAN P, WALTERS P, et al. Deep learning for the life sciences [M]. Sevastopol: O’ Reilly Media, 2019.

[4] HUANG K, FU T, GLASS L M, et al. DeepPurpose: a deep learning library for drug-target interaction prediction [J]. Bioinform, 2020, 36: 5545-5547.

[5] KORSHUNOVA M, GINSBURG B, TROPSHA A, et al. OpenChem: a deep learning toolkit for computational chemistry and drug design [J]. J Chem Inf Model, 2021, 61: 7-13.

[6] WANG J-C, CHU P-Y, CHEN C-M, et al. IdTarget: a web server for identifying protein targets of small chemical molecules with robust scoring functions and a divide-and-conquer docking approach [J]. Nucleic Acids Res, 2012, 40: 393-399.

[7] LI H, GAO Z, KANG L, et al. TarFisDock: a web server for identifying drug targets with docking approach [J]. Nucleic Acids Res, 2006, 34: 219-224.

[8] GONG J, CAI C, LIU X, et al. ChemMapper: a versatile web server for exploring pharmacology and chemical structure association based on molecular 3D similarity method [J]. Bioinform, 2013, 29: 1827-1829.

[9] LIU X, OUYANG S, YU B, et al. PharmMapper server: a web server for potential drug target identification using pharmacophore mapping approach [J]. Nucleic Acids Res, 2010, 38: 609-614.

[10] GFELLE R D , GROSDIDE R A , WIRT H M , e t a l . SwissTargetPrediction: a web server for target prediction of bioactive small molecules [J]. Nucleic Acids Res, 2014, 42: 32-38.

[11] BAI F, MORCOS F, CHENG R R, et al. Elucidating the druggable interface of protein-protein interactions using fragment docking and coevolutionary analysis [J]. Pro Natl Acad Sci USA, 2016, 113: 8051-8058.

[12] ZENG X, ZHU S, LU W, et al. Target identification among known drugs by deep learning from heterogeneous networks [J]. Chem Sci, 2020, 11: 1775-1797.

[13] CONG Q, ANISHCHENKO I, OVCHINNIKOV S, et al. Protein interaction networks revealed by proteome coevolution [J]. Science, 2019, 365: 185-189.

[14] HASHEMIFAR S, NEYSHABUR B, KHAN A A, et al. Predicting protein-protein interactions through sequence-based deep learning [J]. Bioinform, 2018, 34: 802-810.

[15] GAINZA P, SVERRISSON F, MONTI F, et al. Deciphering interaction fingerprints from protein molecular surfaces using geometric deep learning [J]. Nat Methods, 2020, 17: 184-192.

[16] ALTSCHUL S F, GISH W, MILLER W, et al. Basic local alignment search tool [J]. J Mol Biol, 1990, 215: 403-410.

[17] ALTSCHUL S F, MADDEN T L, SCHÄFFER A A, et al. Gapped BLAST and PSI-BLAST: a new generation of protein database search programs [J]. Nucleic Acids Res, 1997, 25: 3389-3402.

[18] FINN R D, CLEMENTS J, EDDY S R. HMMER web server: interactive sequence similarity searching [J]. Nucleic Acids Res, 2011, 39: 29-37.

[19] DAMIAN S, ANDREA F, STEFAN W, et al. STRING v10: protein– protein interaction networks, integrated over the tree of life [J]. Nucleic Acids Res, 2015, 43: D1.

[20] KULMANOV M, KHAN M A, HOEHNDORF R. DeepGO: Predicting protein functions from sequence and interactions using a deep ontologyaware classifier [J]. Bioinform, 2017, 34(4): 660-668.

[21] CAO Y, SHEN Y. TALE: transformer-based protein function annotation with joint sequence-label embedding [J]. Bioinform, 2021, 37(18): 2825-2833.

[22] MA J, SHERIDAN R P, LIAW A, et al. Deep neural nets as a method for quantitative structure-activity relationships [J]. J Chem Info Model, 2015, 55: 263-274.

[23] XU Y, MA J, LIAW A, et al. Demystifying multitask deep neural networks for quantitative structure-activity relationships [J]. J Chem Info Model, 2017, 57: 2490-2504.

[24] SENIOR A W, EVANS R, JUMPER J, et al. Improved protein structure prediction using potentials from deep learning [J]. Nature, 2020, 577: 706-710.

[25] ROHL C A, STRAUSS C E M, MISURA K M S, et al. Protein structure prediction using rosetta [J]. Meth Enzymol, 2004, 383: 66-93.

[26] YANG J, ANISHCHENKO I, PARK H, et al. Improved protein structure prediction using predicted interresidue orientations [J]. Pro Natl Acad Sci USA, 2020, 117: 1496-1503.

[27] CALLAWAY E. ‘It will change everything’: deepmind’s AI makes gigantic leap in solving protein structures [J]. Nature, 2020, 588: 203-204.

[28] SATO K, AKIYAMA M, SAKAKIBARA Y. RNA secondary structure prediction using deep learning with thermodynamic integration [J]. Nature Comm, 2021, 12: 941.

[29] SINGH J, HANSON J, PALIWAL K, et al. RNA secondary structure prediction using an ensemble of two-dimensional deep neural networks and transfer learning [J]. Nature Comm, 2019, 10: 5407.

[30] ZHANG H, ZHANG L, MATHEWS D H, et al. LinearPartition: linear-time approximation of RNA folding partition function and base-pairing probabilities [J]. Bioinform, 2020, 36: 258-267.

[31] ZHANG T, SINGH J, LITFIN T, et al. RNAcmap: a fully automatic method for predicting contact maps of RNAs by evolutionary coupling analysis [J]. Bioinform, 2021. DOI: 10.1093/bioinformatics/ btab391.

[32] LI J, ZHU W, WANG J, et al. RNA3DCNN: local and global quality assessments of RNA 3D structures using 3D deep convolutional neural networks [J]. PLoS Comp Biol, 2018, 14: e1006514.

[33] NGAN C H, BOHNUUD T, MOTTARELLA S E, et al. FTMAP: extended protein mapping with user-selected probe molecules [J]. Nucleic Acids Res, 2012, 40: W271-W275.

[34] KOZLOVSKII I, POPOV P. Spatiotemporal identification of druggable binding sites using deep learning [J]. Commun Biol, 2020, 3: 1-12.

[35] LI Z, YAN X, WEI Q, et al. PointSite: a point cloud segmentation tool for identification of protein ligand binding atoms [J]. BioRxiv, 2019: 831131. DOI: 10.1101/83.1131.

[36] ÖZTÜRK H, ÖZGÜR A, OZKIRIMLI E. DeepDTA: deep drugtarget binding affinity prediction [J]. Bioinform, 2018, 34: 821-829.

[37] ZHENG L, FAN J, MU Y. OnionNet: a multiple-layer intermolecular-contact-based convolutional neural network for protein-ligand binding affinity prediction [J]. ACS Omega, 2019, 4: 15956-15965.

[38] JIMÉNEZ J, ŠKALIČ M, MARTÍNEZ-ROSELL G, et al. KDEEP: protein-ligand absolute binding affinity prediction via 3D-convolutional neural networks [J]. J Chem Info Model, 2018, 58: 287-296.

[39] HASSAN-HARRIROU H, ZHANG C, LEMMIN T. RosENet: improving binding affinity prediction by leveraging molecular mechanics energies with an ensemble of 3D convolutional neural networks [J]. J Chem Info Model, 2020, 60: 2791-2802.

[40] LIN X, ZHAO K, XIAO T, et al. DeepGS: deep representation learning of graphs and sequences for drug-target binding affinity prediction [J]. ECAI, 2020: 1301-1308.

[41] MÉNDEZ-LUCIO O, BAILLIF B, CLEVERT D-A, et al. De novo generation of hit-like molecules from gene expression signatures using artificial intelligence [J]. Nature Comm, 2020, 11: 10.

[42] GOODFELLOW I J, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial networks [J]. Commun ACM, 2020, 63: 139-144.

[43] POPOVA M, ISAYEV O, TROPSHA A. Deep reinforcement learning for de novo drug design [J]. Sci Adv, 2018, 4: eaap7885. [44] KHEMCHANDANI Y, O’HAGAN S, SAMANTA S, et al. DeepGraphMolGen, a multi-objective, computational strategy for generating molecules with desirable properties: a graph convolution and reinforcement learning approach [J]. J Chem , 2020, 12: 53.

[45] LIU B, RAMSUNDAR B, KAWTHEKAR P, et al. Retrosynthetic reaction prediction using neural sequence-to-sequence models [J]. ACS Cent Sci, 2017, 3: 1103-1113.

[46] SHI C, XU M, GUO H, et al. A graph to graphs framework for retrosynthesis prediction [J]. ICML, 2020,119: 8818-8827. [47] CHENG F, LI W, ZHOU Y, et al. AdmetSAR: A comprehensive source and free tool for assessment of chemical admet properties [J]. J Chem Info Model, 2012, 52: 3099-3105.

[48] YANG H, LOU C, SUN L, et al. admetSAR 2.0: web-service for prediction and optimization of chemical ADMET properties [J]. Bioinform, 2019, 35: 1067-1069.

[49] LIU K, SUN X, JIA L, et al. Chemi-net: a graph convolutional network for accurate drug property prediction [J]. Int J Mol Sci, 2019, 20: 3389.

[50] WENZEL J, MATTER H, SCHMIDT F. Predictive multitask deep neural network models for adme-tox properties: learning from large data sets [J]. J Chem Info Model, 2019, 59: 1253-1268.