人工智能驱动的分子生成方法与数据资源综述

doi:10.11871/jfdc.issn.2096-742X.2026.02.014

数据与计算发展前沿 ›› 2026, Vol. 8 ›› Issue (2): 184-203.

CSTR: 32002.14.jfdc.CN10-1649/TP.2026.02.014

doi: 10.11871/jfdc.issn.2096-742X.2026.02.014

人工智能驱动的分子生成方法与数据资源综述

许黄超^1,²(),张宝花^1,^*(),刘倩¹,金钟^1,^*()

¹ 中国科学院计算机网络信息中心，北京 100083
² 中国科学院大学，北京 100049

收稿日期:2025-08-08 出版日期:2026-04-20 发布日期:2026-04-23
通讯作者: *张宝花（E-mail: zhangbh@cnic.cn）；金钟 (E-mail: zjin@sccas.cn)
作者简介:许黄超，中国科学院计算机网络信息中心，中国科学院大学，博士研究生，主要研究方向为人工智能辅助药物设计。
本文承担工作为：文献的收集整理与整体内容的撰写。
XU Huangchao is a Ph.D. student at the Computer Network Information Center, Chinese Academy of Science, and the University of Chinese Academy of Science. Her main research interests include AI-assisted drug design.
In this paper, she is responsible for literature collection and organization, as well as paper writing.
E-mail: hcxu@cnic.cn|张宝花，中国科学院计算机网络信息中心，高级工程师，主要研究方向为超智融合技术及在药物等领域应用。
本文承担工作为：框架设计与指导。
ZHANG Baohua is a senior engineer at the Computer Network Information Center, Chinese Academy of Science. Her main interests include HPC&AI integration technology and its application in fields such as pharmaceuticals.
In this work, she is responsible for framework design and guidance.
E-mail: zhangbh@cnic.cn|金钟，中国科学院计算机网络信息中心，研究员，主要研究领域为生物医学计算与并行软件实现。
本文承担工作为：调研方向和内容指导。
JIN Zhong is a professor at the Computer Network Information Center, Chinese Academy of Science. His main research areas include biomedical computing and parallel software implementation.
In this paper, he is responsible for guiding the research direction and content.
E-mail: zjin@sccas.cn
基金资助:
四大慢病重大专项(2024ZD0533504)

A Review on AI-Driven Methods and Data Resources for Molecule Generation

XU Huangchao^1,²(),ZHANG Baohua^1,^*(),LIU Qian¹,JIN Zhong^1,^*()

¹ Computer Network Information Center, Chinese Academy of Sciences, Beijing 100083, China
² University of Chinese Academy of Sciences, Beijing 100049, China

Received:2025-08-08 Online:2026-04-20 Published:2026-04-23

摘要/Abstract

摘要：

【目的】 在人工智能（AI）技术与海量分子数据的双重驱动下，AI赋能的分子生成已成为药物设计与化学创新的关键技术。本文聚焦于小分子设计，旨在系统综述AI驱动的小分子生成方法及在药物研发中的应用。【文献范围】梳理了国内外支持小分子生成的主要数据资源、生成方法与应用研究。【方法】 围绕变分自编码器、生成对抗网络、Transformer、扩散模型及大语言模型等技术路线，介绍当前主流模型及其核心机制，结合靶点引导、结构约束与语言建模等策略展开归纳。【结果】 AI小分子生成方法在多个应用场景中展现出显著优势，但对数据质量、算法复杂度和计算资源提出更高要求。【局限】受篇幅限制，本文未能全面涵盖该领域的所有分支和最新进展。【结论】 AI驱动的小分子生成方法正在加速分子发现和药物设计创新进程，构建AI-ready的高质量数据集，提升小分子生成模型的可控性和泛化能力，完善生成评价体系将是未来重要的研究方向。

关键词: 药物发现, 小分子生成, 深度生成模型, 人工智能

Abstract:

[Objective] Advances in artificial intelligence (AI) and the rapid growth of molecular data have made AI-driven molecular generation a key technology in drug design and chemical innovation. This review focuses on small molecule design, summarizing AI-driven generation methods, key data resources, and their current applications in drug discovery. [Coverage] We survey key data resources that support small molecule generation, summarize representative generative modeling approaches, and analyze recent applications studies from both domestic and international research. [Methods] The article reviews mainstream deep learning architectures used in molecular generation, including Variational Autoencoders (VAEs), Generative Adversarial Networks (GANs), Transformers, Diffusion Models, and Large Language Models (LLMs). It further summarizes advanced strategies including target-guided, fragment-based, and language model-driven generation. [Results] AI-driven small molecule generation has shown strong effectiveness in various applications. Growing model complexity and data volume require higher-quality datasets, advanced algorithms, and increased computational resources. [Limitations] This review mainly covers key data resources, core methods, and main applications in small molecule generation, without encompassing all subfields or the latest developments in this rapidly evolving area. [Conclusions] The development of domain-specific data resources has enabled a range of AI-driven small molecule generation methods, accelerating molecular discovery and drug design. Building high-quality, AI-ready datasets, improving model controllability and generalization, and refining evaluation systems remain important directions for future research.

Key words: drug design, molecule generation, deep generative models, artificial intelligence

许黄超, 张宝花, 刘倩, 金钟. 人工智能驱动的分子生成方法与数据资源综述[J]. 数据与计算发展前沿, 2026, 8(2): 184-203.

XU Huangchao, ZHANG Baohua, LIU Qian, JIN Zhong. A Review on AI-Driven Methods and Data Resources for Molecule Generation[J]. Frontiers of Data and Computing, 2026, 8(2): 184-203, https://cstr.cn/32002.14.jfdc.CN10-1649/TP.2026.02.014.

图/表 4

图1

表1

图2

图3

参考文献 112

[1]	JAYATUNGA M, AYERS M, BRUENS L, et al. How successful are AI-discovered drugs in clinical trials? A first analysis and emerging lessons[J]. Drug Discovery Today, 2024, 29: 104009. doi: 10.1016/j.drudis.2024.104009
[2]	IRWIN J J, TANG K G, YOUNG J, et al. ZINC20—A Free Ultralarge-Scale Chemical Database for Ligand Discovery[J]. Journal of Chemical Information and Modeling, 2020, 60(12): 6065-6073. doi: 10.1021/acs.jcim.0c00675
[3]	TINGLE B I, TANG K G, CASTANON M, et al. ZINC-22─A Free Multi-Billion-Scale Database of Tangible Compounds for Ligand Discovery[J]. Journal of Chemical Information and Modeling, 2023, 63(4): 1166-1176. doi: 10.1021/acs.jcim.2c01253 pmid: 36790087
[4]	GAULTON A, BELLIS L J, BENTO A P, et al. ChEMBL: a large-scale bioactivity database for drug discovery[J]. Nucleic Acids Research, 2012, 40(Database issue): D1100-D1107. doi: 10.1093/nar/gkr777
[5]	ZDRAZIL B, FELIX E, HUNTER F, et al. The ChEMBL Database in 2023: a drug discovery platform spanning multiple bioactivity data types and time periods[J]. Nucleic Acids Research, 2024, 52(D1): D1180-D1192. doi: 10.1093/nar/gkad1004
[6]	KIM S, THIESSEN P A, BOLTON E E, et al. PubChem Substance and Compound databases[J]. Nucleic Acids Research, 2016, 44(Database issue): D1202-D1213. doi: 10.1093/nar/gkv951
[7]	KIM S, CHEN J, CHENG T, et al. PubChem 2023 update[J]. Nucleic Acids Research, 2023, 51(D1): D1373-D1380. doi: 10.1093/nar/gkac956
[8]	GRYGORENKO O O, RADCHENKO D S, DZIUBA I, et al. Generating Multibillion Chemical Space of Readily Accessible Screening Compounds[J]. iScience, 2020, 23(11): 101681. doi: 10.1016/j.isci.2020.101681
[9]	WISHART D S, FEUNANG Y D, GUO A C, et al. DrugBank 5.0: a major update to the DrugBank database for 2018[J]. Nucleic Acids Research, 2018, 46(D1): D1074-D1082. doi: 10.1093/nar/gkx1037
[10]	KNOX C, WILSON M, KLINGER C M, et al. DrugBank 6.0: the DrugBank Knowledgebase for 2024[J]. Nucleic Acids Research, 2024, 52(D1): D1265-D1275. doi: 10.1093/nar/gkad976
[11]	SHANG J, SUN H, LIU H, et al. Comparative analyses of structural features and scaffold diversity for purchasable compound libraries[J]. Journal of Cheminformatics, 2017, 9(1): 25. doi: 10.1186/s13321-017-0212-4 pmid: 29086044
[12]	BURLEY S K, BERMAN H M, BHIKADIYA C, et al. RCSB Protein Data Bank: biological macromolecular structures enabling research and education in fundamental biology, biomedicine, biotechnology and energy[J]. Nucleic Acids Research, 2019, 47(D1): D464-D474. doi: 10.1093/nar/gky1004
[13]	UNIPROT CONSORTIUM T, BATEMAN A, MARTIN M J, et al. UniProt: the universal protein knowledgebase[J]. Nucleic Acids Research, 2017, 45(D1): D158-D169. doi: 10.1093/nar/gkw1099
[14]	CONSORTIUM T U, BATEMAN A, MARTIN M J, et al. UniProt: the Universal Protein Knowledgebase in 2023[J]. Nucleic Acids Research, 2023, 51(D1): D523-D531. doi: 10.1093/nar/gkac1052
[15]	JUMPER J, EVANS R, PRITZEL A, et al. Highly accurate protein structure prediction with AlphaFold[J]. Nature, 2021, 596(7873): 583-589. doi: 10.1038/s41586-021-03819-2
[16]	ABRAMSON J, ADLER J, DUNGER J, et al. Accurate structure prediction of biomolecular interactions with AlphaFold 3[J]. Nature, 2024, 630(8016): 493-500. doi: 10.1038/s41586-024-07487-w
[17]	WANG R, FANG X, LU Y, et al. The PDBbind database: methodologies and updates[J]. Journal of Medicinal Chemistry, 2005, 48(12): 4111-4119. pmid: 15943484
[18]	LIU Z, LI Y, HAN L, et al. PDB-wide collection of binding data: current status of the PDBbind database[J]. Bioinformatics, 2015, 31(3): 405-412. doi: 10.1093/bioinformatics/btu626 pmid: 25301850
[19]	FRANCOEUR P G, MASUDA T, SUNSERI J, et al. Three-Dimensional Convolutional Neural Networks and a Cross-Docked Data Set for Structure-Based Drug Design[J]. Journal of Chemical Information and Modeling, 2020, 60(9): 4200-4215. doi: 10.1021/acs.jcim.0c00411 pmid: 32865404
[20]	VAN LINDEN O P J, KOOISTRA A J, LEURS R, et al. KLIFS: A Knowledge-Based Structural Database To Navigate Kinase-Ligand Interaction Space[J]. Journal of Medicinal Chemistry, 2014, 57(2): 249-277. doi: 10.1021/jm400378w pmid: 23941661
[21]	KANEV G K, DE GRAAF C, WESTERMAN B A, et al. KLIFS: an overhaul after the first 5 years of supporting kinase research[J]. Nucleic Acids Research, 2020, 49(D1): D562-D569. doi: 10.1093/nar/gkaa895
[22]	GILSON M K, LIU T, BAITALUK M, et al. BindingDB in 2015: A public database for medicinal chemistry, computational chemistry and systems pharmacology[J]. Nucleic Acids Research, 2016, 44(Database issue): D1045-D1053. doi: 10.1093/nar/gkv1072
[23]	LIU T, HWANG L, BURLEY S K, et al. BindingDB in 2024: a FAIR knowledgebase of protein-small molecule binding data[J]. Nucleic Acids Research, 2025, 53(D1): D1633-D1644. doi: 10.1093/nar/gkae1075 pmid: 39574417
[24]	WU Z, RAMSUNDAR B, FEINBERG E N, et al. MoleculeNet: A Benchmark for Molecular Machine Learning[Z]. arXiv, 2018[2025-07-21]. http://arxiv.org/abs/1703.00564.
[25]	POLYKOVSKIY D, ZHEBRAK A, SANCHEZ-LENGELING B, et al. Molecular Sets (MOSES): A Benchmarking Platform for Molecular Generation Models[Z]. arXiv, 2020[2025-07-10]. http://arxiv.org/abs/1811.12823.
[26]	BROWN N, FISCATO M, SEGLER M H S, et al. GuacaMol: Benchmarking Models for de Novo Molecular Design[J]. Journal of Chemical Information and Modeling, 2019, 59(3): 1096-1108. doi: 10.1021/acs.jcim.8b00839 pmid: 30887799
[27]	HUANG K, FU T, GAO W, et al. Therapeutics Data Commons: Machine Learning Datasets and Tasks for Drug Discovery and Development[Z]. arXiv, 2021[2025-07-10]. http://arxiv.org/abs/2102.09548.
[28]	DONG J, WANG N N, YAO Z J, et al. ADMETlab: a platform for systematic ADMET evaluation based on a comprehensively collected ADMET database[J]. Journal of Cheminformatics, 2018, 10(1): 29. doi: 10.1186/s13321-018-0283-x pmid: 29943074
[29]	XIONG G, WU Z, YI J, et al. ADMETlab 2.0: an integrated online platform for accurate and comprehensive predictions of ADMET properties[J]. Nucleic Acids Research, 2021, 49(W1): W5-W14. doi: 10.1093/nar/gkab255 pmid: 33893803
[30]	FU L, SHI S, YI J, et al. ADMETlab 3.0: an updated comprehensive online ADMET prediction platform enhanced with broader coverage, improved performance, API functionality and decision support[J]. Nucleic Acids Research, 2024, 52(W1): W422-W431. doi: 10.1093/nar/gkae236
[31]	ROGERS D, HAHN M. Extended-Connectivity Fingerprints[J]. Journal of Chemical Information and Modeling, 2010, 50(5): 742-754. doi: 10.1021/ci100050t pmid: 20426451
[32]	KUWAHARA H, GAO X. Analysis of the effects of related fingerprints on molecular similarity using an eigenvalue entropy approach[J]. Journal of Cheminformatics, 2021, 13(1): 27. doi: 10.1186/s13321-021-00506-2 pmid: 33757582
[33]	JAEGER S, FULLE S, TURK S. Mol2vec: unsupervised machine learning approach with chemical intuition[J]. Journal of chemical information and modeling, 2018, 58(1): 27-35. doi: 10.1021/acs.jcim.7b00616 pmid: 29268609
[34]	FABIAN B, EDLICH T, GASPAR H, et al. Molecular representation learning with language models and domain-relevant auxiliary tasks[Z]. arXiv, 2020[2025-07-21]. http://arxiv.org/abs/2011.13230.
[35]	XIONG Z, WANG D, LIU X, et al. Pushing the Boundaries of Molecular Representation for Drug Discovery with the Graph Attention Mechanism[J]. Journal of Medicinal Chemistry, 2020, 63(16): 8749-8760. doi: 10.1021/acs.jmedchem.9b00959 pmid: 31408336
[36]	RONG Y, BIAN Y, XU T, et al. Self-Supervised Graph Transformer on Large-Scale Molecular Data[Z]. arXiv, 2020[2025-07-21]. http://arxiv.org/abs/2007.02835.
[37]	LIU S, WANG H, LIU W, et al. Pre-training Molecular Graph Representation with 3D Geometry[Z]. arXiv, 2022[2025-07-21]. http://arxiv.org/abs/2110.07728.
[38]	BATZNER S, MUSAELIAN A, SUN L, et al. E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentials[J/OL]. Nature Communications, 2022, 13(1)[2025-07-10]. http://arxiv.org/abs/2101.03164.
[39]	FUCHS F B, WORRALL D E, FISCHER V, et al. SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networks[Z]. arXiv, 2020[2025-07-10]. http://arxiv.org/abs/2006.10503.
[40]	LU S, GAO Z, HE D, et al. Data-driven quantum chemical property prediction leveraging 3D conformations with Uni-Mol+[J]. Nature Communications, 2024, 15(1): 7104. doi: 10.1038/s41467-024-51321-w pmid: 39160169
[41]	CAI H, ZHANG H, ZHAO D, et al. FP-GNN: a versatile deep learning architecture for enhanced molecular property prediction[J]. Briefings in Bioinformatics, 2022, 23(6): bbac408.
[42]	WU S, YU D, TAN X, et al. CLaMP: Contrastive Language-Music Pre-training for Cross-Modal Symbolic Music Information Retrieval[Z]. arXiv, 2023[2025-07-21]. http://arxiv.org/abs/2304.11029.
[43]	KAUFMAN B, WILLIAMS E, UNDERKOFFLER C, et al. COATI: multi-modal contrastive pre-training for representing and traversing chemical space[Z]. ChemRxiv, 2023[2025-07-21]. https://chemrxiv.org/engage/chemrxiv/article-details/64e8137fdd1a73847f73f7aa.
[44]	LIN Z, AKIN H, RAO R, et al. Evolutionary-scale prediction of atomic level protein structure with a language model[Z]. bioRxiv, 2022: 2022.07.20.500902[2025-07-10]. https://www.biorxiv.org/content/10.1101/2022.07.20.500902v3.
[45]	HAYES T, RAO R, AKIN H, et al. Simulating 500 million years of evolution with a language model[Z]. bioRxiv, 2024: 2024.07.01.600583[2025-07-10]. https://www.biorxiv.org/content/10.1101/2024.07.01.600583v1.
[46]	BRANDES N, OFER D, PELEG Y, et al. ProteinBERT: a universal deep-learning model of protein sequence and function[J]. Bioinformatics, 2022, 38(8): 2102-2110. doi: 10.1093/bioinformatics/btac020 pmid: 35020807
[47]	RAO R, BHATTACHARYA N, THOMAS N, et al. Evaluating Protein Transfer Learning with TAPE[Z]. arXiv, 2019[2025-07-10]. http://arxiv.org/abs/1906.08230.
[48]	JIANG M, LI Z, ZHANG S, et al. Dru-target affinity prediction using graph neural network and contact maps[J]. RSC Advances, 2020, 10(35): 20701-20712. doi: 10.1039/D0RA02297G
[49]	KLOCZKOWSKI A, JERNIGAN R L, WU Z, et al. Distance Matrix-Based Approach to Protein Structure Prediction[J]. Journal of structural and functional genomics, 2009, 10(1): 67-81. doi: 10.1007/s10969-009-9062-2 pmid: 19224393
[50]	ZHAO L, WANG H, SHI S. PocketDTA: an advanced multimodal architecture for enhanced prediction of drug-target affinity from 3D structural data of target binding pockets[J]. Bioinformatics, 2024, 40(10): btae594.
[51]	JING B, EISMANN S, SURIANA P, et al. Learning from Protein Structure with Geometric Vector Perceptrons[Z]. arXiv, 2021[2025-07-10]. http://arxiv.org/abs/2009.01411.
[52]	LIU C, WANG J, CAI Z, et al. Dynamic PDB: A New Dataset and a SE(3) Model Extension by Integrating Dynamic Behaviors and Physical Properties in Protein Structures[Z]. arXiv, 2024[2025-07-21]. http://arxiv.org/abs/2408.12413.
[53]	JIN W, BARZILAY R, JAAKKOLA T. Junction Tree Variational Autoencoder for Molecular Graph Generation[Z]. arXiv, 2019[2025-07-10]. http://arxiv.org/abs/1802.04364.
[54]	KUSNER M J, PAIGE B, HERNÁNDEZ-LOBATO J M. Grammar Variational Autoencoder[C]// Proceedings of the 34th International Conference on Machine Learning. PMLR, 2017: 1945-1954.
[55]	CAO N D, KIPF T. MolGAN: An implicit generative model for small molecular graphs[Z]. arXiv, 2022[2025-07-21]. http://arxiv.org/abs/1805.11973.
[56]	GUIMARAES G L, SANCHEZ-LENGELING B, OUTEIRAL C, et al. Objective-Reinforced Generative Adversarial Networks (ORGAN) for Sequence Generation Models[Z]. arXiv, 2018[2025-07-10]. http://arxiv.org/abs/1705.10843.
[57]	ZANG C, WANG F. MoFlow: An Invertible Flow Model for Generating Molecular Graphs[Z]. arXiv, 2020[2025-07-10]. http://arxiv.org/abs/2006.10137.
[58]	SONG Y, GONG J, XU M, et al. Equivariant Flow Matching with Hybrid Probability Transport[Z]. arXiv, 2023[2025-08-15]. http://arxiv.org/abs/2312.07168.
[59]	XU M, YU L, SONG Y, et al. GeoDiff: a Geometric Diffusion Model for Molecular Conformation Generation[Z]. arXiv, 2022[2025-07-10]. http://arxiv.org/abs/2203.02923.
[60]	HOOGEBOOM E, SATORRAS V G, VIGNAC C, et al. Equivariant Diffusion for Molecule Generation in 3D[Z]. arXiv, 2022[2025-07-10]. http://arxiv.org/abs/2203.17003.
[61]	SHERSTINSKY A. Fundamentals of Recurrent Neural Network (RNN) and Long Short-Term Memory (LSTM) Network[J]. Physica D: Nonlinear Phenomena, 2020, 404: 132306. doi: 10.1016/j.physd.2019.132306
[62]	CHUNG J, GULCEHRE C, CHO K, et al. Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling[Z]. arXiv, 2014[2025-07-10]. http://arxiv.org/abs/1412.3555.
[63]	BLASCHKE T, ARÚS-POUS J, CHEN H, et al. REINVENT 2.0: An AI Tool for De Novo Drug Design[J]. Journal of Chemical Information and Modeling, 2020, 60(12): 5918-5922. doi: 10.1021/acs.jcim.0c00915 pmid: 33118816
[64]	LOEFFLER H H, HE J, TIBO A, et al. Reinvent 4: Modern AI-driven generative molecule design[J]. Journal of Cheminformatics, 2024, 16(1): 20. doi: 10.1186/s13321-024-00812-5 pmid: 38383444
[65]	IRWIN R, DIMITRIADIS S, HE J, et al. Chemformer: a pre-trained transformer for computational chemistry[J]. Machine Learning: Science and Technology, 2022, 3(1): 015022. doi: 10.1088/2632-2153/ac3ffb
[66]	OPENAI, ACHIAM J, ADLER S, et al. GPT-4 Technical Report[Z]. arXiv, 2024[2025-07-13]. http://arxiv.org/abs/2303.08774.
[67]	KEVIAN D, SYED U, GUO X, et al. Capabilities of Large Language Models in Control Engineering: A Benchmark Study on GPT-4, Claude 3 Opus, and Gemini 1.0 Ultra[Z]. arXiv, 2024[2025-07-13]. http://arxiv.org/abs/2404.03647.
[68]	TEAM G, ANIL R, BORGEAUD S, et al. Gemini: A Family of Highly Capable Multimodal Models[Z]. arXiv, 2025[2025-07-13]. http://arxiv.org/abs/2312.11805.
[69]	FREY N C, SOKLASKI R, AXELROD S, et al. Neural scaling of deep chemical models[J]. Nature Machine Intelligence, 2023, 5(11): 1297-1305. doi: 10.1038/s42256-023-00740-3
[70]	BAGAL V, AGGARWAL R, VINOD P K, et al. MolGPT: Molecular Generation Using a Transformer-Decoder Model[J]. Journal of Chemical Information and Modeling, 2022, 62(9): 2064-2076. doi: 10.1021/acs.jcim.1c00600
[71]	LI Y, GAO C, SONG X, et al. DrugGPT: A GPT-based Strategy for Designing Potential Ligands Targeting Specific Proteins[J]. bioRxiv, 2023: 2023.06.29.543848.
[72]	FANG Y, LIANG X, ZHANG N, et al. Mol-Instructions: A Large-Scale Biomolecular Instruction Dataset for Large Language Models[Z]. arXiv, 2024[2025-07-13]. http://arxiv.org/abs/2306.08018.
[73]	LUO Y, YANG K, HONG M, et al. MolFM: A Multimodal Molecular Foundation Model[Z]. arXiv, 2023[2025-07-13]. http://arxiv.org/abs/2307.09484.
[74]	SEGLER M H S, KOGEJ T, TYRCHAN C, et al. Generating Focused Molecule Libraries for Drug Discovery with Recurrent Neural Networks[J]. ACS Central Science, 2018, 4(1): 120-131. doi: 10.1021/acscentsci.7b00512 pmid: 29392184
[75]	HONDA S, SHI S, UEDA H R. SMILES Transformer: Pre-trained Molecular Fingerprint for Low Data Drug Discovery[Z]. arXiv, 2019[2025-07-13]. http://arxiv.org/abs/1911.04738.
[76]	SIMONOVSKY M, KOMODAKIS N. GraphVAE: Towards Generation of Small Graphs Using Variational Autoencoders[Z]. arXiv, 2018[2025-07-13]. http://arxiv.org/abs/1802.03480.
[77]	SHI C, XU M, ZHU Z, et al. GraphAF: a Flow-based Autoregressive Model for Molecular Graph Generation[Z]. arXiv, 2020[2025-07-13]. http://arxiv.org/abs/2001.09382.
[78]	XIANG Y, ZHAO H, MA C, et al. Instruction-Based Molecular Graph Generation with Unified Text-Graph Diffusion Model[Z]. arXiv, 2024[2025-07-21]. http://arxiv.org/abs/2408.09896.
[79]	NI Y, FENG S, CHI H, et al. Straight-Line Diffusion Model for Efficient 3D Molecular Generation[Z]. arXiv, 2025[2025-07-21]. http://arxiv.org/abs/2503.02918.
[80]	KIRCHMEYER M, PINHEIRO P O, SAREMI S. Score-based 3D molecule generation with neural fields[Z]. arXiv, 2025[2025-07-21]. http://arxiv.org/abs/2501.08508.
[81]	ZHANG O, WANG T, WENG G, et al. Learning on topological surface and geometric structure for 3D molecular generation[J]. Nature Computational Science, 2023, 3(10): 849-859. doi: 10.1038/s43588-023-00530-2 pmid: 38177756
[82]	ZHANG O, ZHANG J, JIN J, et al. ResGen is a pocket-aware 3D molecular generation model based on parallel multiscale modelling[J]. Nature Machine Intelligence, 2023, 5(9):1020-1030. doi: 10.1038/s42256-023-00712-7
[83]	SCHNEUING A, HARRIS C, DU Y, et al. Structure-based Drug Design with Equivariant Diffusion Models[Z]. arXiv, 2024[2025-07-13]. http://arxiv.org/abs/2210.13695.
[84]	GUAN J, QIAN W W, PENG X, et al. 3D Equivariant Diffusion for Target-Aware Molecule Generation and Affinity Prediction[Z]. arXiv, 2023[2025-07-13]. http://arxiv.org/abs/2303.03543.
[85]	GUAN J, ZHOU X, YANG Y, et al. DecompDiff: Diffusion Models with Decomposed Priors for Structure-Based Drug Design[Z]. arXiv, 2024[2025-07-13]. http://arxiv.org/abs/2403.07902.
[86]	QIAN H, HUANG W, TU S, et al. KGDiff: towards explainable target-aware molecule generation with knowledge guidance[J]. Briefings in Bioinformatics, 2023, 25(1): bbad435.
[87]	DU H, JIANG D, ZHANG O, et al. A flexible data-free framework for structure-based de novo drug design with reinforcement learning[J]. Chemical Science, 2023, 14(43): 12166-12181. doi: 10.1039/D3SC04091G
[88]	ZHANG O, HUANG Y, CHENG S, et al. FragGen: towards 3D geometry reliable fragment-based molecular generation[J]. Chemical Science, 2024, 15(46): 19452-19465. doi: 10.1039/d4sc04620j pmid: 39568888
[89]	DIAO Y, HU F, SHEN Z, et al. MacFrag: segmenting large-scale molecules to obtain diverse fragments with high qualities[J]. Bioinformatics, 2023, 39(1): btad012.
[90]	SYDOW D, SCHMIEL P, MORTIER J, et al. KinFragLib: Exploring the Kinase Inhibitor Space Using Subpocket-Focused Fragmentation and Recombination[J]. Journal of Chemical Information and Modeling, 2020, 60(12): 6081-6094. doi: 10.1021/acs.jcim.0c00839 pmid: 33155465
[91]	LEE S, KREIS K, VECCHAM S P, et al. Molecule Generation with Fragment Retrieval Augmentation[Z]. arXiv, 2024[2025-07-13]. http://arxiv.org/abs/2411.12078.
[92]	IMRIE F, HADFIELD T E, BRADLEY A R, et al. Deep generative design with 3D pharmacophoric constraints[J]. Chemical Science, 2021, 12(43): 14577-14589. doi: 10.1039/d1sc02436a pmid: 34881010
[93]	ZHU H, ZHOU R, CAO D, et al. A pharmacophore-guided deep learning approach for bioactive molecular generation[J]. Nature Communications, 2023, 14(1): 6234. doi: 10.1038/s41467-023-41454-9 pmid: 37803000
[94]	XIE W, ZHANG J, XIE Q, et al. Accelerating Discovery of Novel and Bioactive Ligands With Pharmacophore-Informed Generative Models[Z]. arXiv, 2024[2025-07-13]. http://arxiv.org/abs/2401.01059.
[95]	ALAKHDAR A, POCZOS B, WASHBURN N. Pharmacophore-Conditioned Diffusion Model for Ligand-Based De Novo Drug Design[Z]. arXiv, 2025[2025-07-21]. http://arxiv.org/abs/2505.10545.
[96]	SUN M, XING J, MENG H, et al. MolSearch: Search-based Multi-objective Molecular Generation and Property Optimization[C]// Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. Washington DC USA: ACM, 2022: 4724-4732.
[97]	LIU Y, ZHU Y, WANG J, et al. A Multi-Objective Molecular Generation Method Based on Pareto Algorithm and Monte Carlo Tree Search[J]. Advanced Science, 2025, 12(20): 2410640. doi: 10.1002/advs.v12.20
[98]	CHEN S. Structure-aware dual-target drug design through collaborative learning of pharmacophore combination and molecular simulation[J]. Chemical Science, 2024, 15(27): 10366-10380. doi: 10.1039/d4sc00094c pmid: 38994407
[99]	BHATTACHARYA D, CASSADY H J, HICKNER M A, et al. Large Language Models as Molecular Design Engines[J]. Journal of Chemical Information and Modeling, 2024, 64(18): 7086-7096. doi: 10.1021/acs.jcim.4c01396 pmid: 39231030
[100]	ZENG Z, YIN B, WANG S, et al. Interactive Molecular Discovery with Natural Language[Z]. arXiv, 2023[2025-07-14]. http://arxiv.org/abs/2306.11976.
[101]	ISHIDA S, SATO T, HONMA T, et al. Large language models open new way of AI-assisted molecule design for chemists[J]. Journal of Cheminformatics, 2025, 17(1): 36. doi: 10.1186/s13321-025-00984-8 pmid: 40128788
[102]	MALIKUSSAID, NUHA H H. VALID-Mol: a Systematic Framework for Validated LLM-Assisted Molecular Design[Z]. arXiv, 2025[2025-07-21]. http://arxiv.org/abs/2506.23339.
[103]	MA P, CHENG Z, CHENG Z, et al. Discovery of EP4 antagonists with image-guided explainable deep learning workflow[J]. National Science Open, 2025, 4(4): 20240015. doi: 10.1360/nso/20240015
[104]	GAO B, HUANG Y, LIU Y, et al. PharmAgents: Building a Virtual Pharma with Large Language Model Agents[Z]. arXiv, 2025[2025-07-14]. http://arxiv.org/abs/2503.22164.
[105]	BICKERTON G R, PAOLINI G V, BESNARD J, et al. Quantifying the chemical beauty of drugs[J]. Nature Chemistry, 2012, 4(2): 90-98. doi: 10.1038/nchem.1243 pmid: 22270643
[106]	ERTL P, SCHUFFENHAUER A. Estimation of synthetic accessibility score of drug-like molecules based on molecular complexity and fragment contributions[J]. Journal of Cheminformatics, 2009, 1(1): 8. doi: 10.1186/1758-2946-1-8 pmid: 20298526
[107]	TROTT O, OLSON A J. AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization and multithreading[J]. Journal of computational chemistry, 2010, 31(2): 455-461. doi: 10.1002/jcc.v31:2
[108]	FRIESNER R A, BANKS J L, MURPHY R B, et al. Glide: A New Approach for Rapid, Accurate Docking and Scoring. 1. Method and Assessment of Docking Accuracy[J]. Journal of Medicinal Chemistry, 2004, 47(7): 1739-1749. doi: 10.1021/jm0306430
[109]	SHEN C, ZHANG X, DENG Y, et al. Boosting Protein-Ligand Binding Pose Prediction and Virtual Screening Based on Residue-Atom Distance Likelihood Potential and Graph Transformer[J]. Journal of Medicinal Chemistry, 2022, 65(15): 10691-10706. doi: 10.1021/acs.jmedchem.2c00991 pmid: 35917397
[110]	ÖZTÜRK H, ÖZGÜR A, OZKIRIMLI E. DeepDTA: deep drug-target binding affinity prediction[J]. Bioinformatics, 2018, 34(17): i821-i829.
[111]	DAINA A, MICHIELIN O, ZOETE V. SwissADME: a free web tool to evaluate pharmacokinetics, drug-likeness and medicinal chemistry friendliness of small molecules[J]. Scientific Reports, 2017, 7(1): 42717. doi: 10.1038/srep42717
[112]	BUTTENSCHOEN M, MORRIS G M, DEANE C M. PoseBusters: AI-based docking methods fail to generate physically valid poses or generalise to novel sequences[Z]. arXiv, 2023[2024-12-09]. http://arxiv.org/abs/2308.05777.

分类	数据集	数据规模	原始数据格式	访问途径
小分子数据	ZINC	37 B	SMILES、SDF	https://cartblanche22.docking.org/
	ChEMBL	22.7 M	SMILES、InChI	httpss://www.ebi.ac.uk/chembl
	PubChem	750 M	SMILES、SDF	httpss://pubchem.ncbi.nlm.nih.gov
	Enamine REAL	10 B	SMILES、SDF	httpss://www.enamine-genez.com/compound_library/real-database
	DrugBank	1.5 M	SMILES、MOL、SDF	httpss://go.drugbank.com
	ChemDiv	1.6 M	SMILES、SDF	httpss://www.chemdiv.com
蛋白质和复合物数据	RCSB PDB	238 K	Experimental PDB	httpss://www.rcsb.org
	AlphaFoldDB	200 M	Predicted PDB	httpss://alphafold.ebi.ac.uk
	UniProt	250 M	Sequence Entry	httpss://www.uniprot.org
	PDBBind	35 K	PDB、SDF	httpss://www.pdbbind-plus.org.cn
	Crossdocked	22.5 M	PDB、SDF	httpss://github.com/gnina/models/tree/master/data/CrossDocked2020
	KLIFS	6 K	PDB、SDF	httpss://klifs.net
	BindingDB	1.2 M	SMILES、Sequence	httpss://www.bindingdb.org
基准评估数据	MoleculeNet	700 K	SMILES	httpss://moleculenet.org
	MOSES	1.9 M	SMILES	httpss://github.com/molecularsets/moses
	GuacaMol	/	SMILES、MD5	httpss://github.com/BenevolentAI/guacamol
	TDC	/	SMILES、PDB	httpss://tdcommons.ai
	ADMET Lab	400 K	SMILES	httpss://admetlab3.scbdd.com

人工智能驱动的分子生成方法与数据资源综述

A Review on AI-Driven Methods and Data Resources for Molecule Generation

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 4

参考文献 112

相关文章 15

编辑推荐

Metrics

本文评价

[1]	何邦彦, 李琦, 孙哲南, 王蕊. 探索基于3D变换的后处理对3D对抗性点云迁移性的影响[J]. 数据与计算发展前沿, 2026, 8(2): 123-140.
[2]	张耀南, 刘景琦, 康建芳, 南卓铜, 田文彪, 敏玉芳, 赵书萍, 王保得. 冰冻圈“大数据+AI+模型”耦合研究范式探索[J]. 数据与计算发展前沿, 2026, 8(2): 3-14.
[3]	刘景琦, 张耀南, 康建芳, 刘杰, 杨治纬, 张智星, 王保得. 基于多智能体协同的高寒山区道路结冰数据工程预警[J]. 数据与计算发展前沿, 2026, 8(2): 40-53.
[4]	张森, 褚铮, 黄松睿, 孟哲令. 中小学人工智能教育综述[J]. 数据与计算发展前沿, 2025, 7(6): 170-178.
[5]	陈长松, 吴跃顺, 梅广. 人工智能安全防护体系的层次化模型研究[J]. 数据与计算发展前沿, 2025, 7(6): 68-76.
[6]	郭惠婕,周泳杰,许建真. 基于多模态人工智能数据融合的中药药动学研究进展[J]. 数据与计算发展前沿, 2025, 7(2): 149-160.
[7]	贾子昂. 基于多源半监督学习的牙齿结构分割[J]. 数据与计算发展前沿, 2025, 7(2): 175-185.
[8]	蔡华谦, 刘逸豪, 关天鹏, 吴恺东, 杨婧如, 罗超然, 朱小杰, 刘佳, 黄罡. DPML: 一种面向科学数据语用的标记语言[J]. 数据与计算发展前沿, 2024, 6(4): 46-58.
[9]	王志永, 刘晶晶, 王新明, 陈博文, 聂伟, 张瀚林, 刘洪海. 孤独症人工智能诊疗进展及前沿[J]. 数据与计算发展前沿, 2024, 6(3): 15-27.
[10]	寇大治. 基于深度学习的口腔全景片牙齿自动分割方法[J]. 数据与计算发展前沿, 2024, 6(3): 162-172.
[11]	何睿琳, 杨欣怡, 孙洪赞, 李晨. 基于图特征的组织病理学图像分析方法的最新发展情况与展望[J]. 数据与计算发展前沿, 2024, 6(2): 101-116.
[12]	朱明明, 曹无敌, 吴林, 王自溪, 廖琦, 张思, 唐晓, 李杰, 王婧, 王彦棡, 王自发. 基于人工智能与大数据的双碳大气环境信息化应用进展与展望[J]. 数据与计算发展前沿, 2023, 5(3): 2-12.
[13]	胡晓彦,徐寄遥,邹自明. “大数据&人工智能”驱动的空间天气科研范式变革初步探索[J]. 数据与计算发展前沿, 2023, 5(2): 24-36.
[14]	齐法制,李刚,李纯,汪璐,张一,张正德,陈刚,罗武鸣,赵丽娜,胡誉,袁野. 基于人工智能的高能物理大数据技术与应用[J]. 数据与计算发展前沿, 2023, 5(2): 50-59.
[15]	王凡,冯立强,曹荣强. 大数据驱动的海洋人工智能服务平台设计与应用[J]. 数据与计算发展前沿, 2023, 5(2): 73-85.