分子药物设计基准测试
molecular drug design benchmarking.
作者信息
Grant Lauren L, Sit Clarissa S
机构信息
Saint Mary's University Halifax NS Canada
出版信息
RSC Med Chem. 2021 Jun 3;12(8):1273-1280. doi: 10.1039/d1md00074h. eCollection 2021 Aug 18.
Abstract
molecular design for drug discovery is a growing field. Deep neural networks (DNNs) are becoming more widespread in their use for machine learning models. As more DNN models are proposed for molecular design, benchmarking methods are crucial for the comparision and validation of these models. This review looks at recently proposed benchmarking methods Fréchet ChemNet Distance, GuacaMol and Molecular Sets (MOSES), and provides a commentary on their future potential applications in molecular drug design and possible next steps for further validation of these benchmarking methods.
摘要
用于药物发现的分子设计是一个不断发展的领域。深度神经网络（DNN）在机器学习模型中的应用越来越广泛。随着越来越多的DNN模型被提出用于分子设计，基准测试方法对于这些模型的比较和验证至关重要。本综述探讨了最近提出的基准测试方法——弗雷歇化学网络距离、瓜卡莫尔和分子集（MOSES），并对它们在分子药物设计中的未来潜在应用以及这些基准测试方法进一步验证的可能后续步骤进行了评论。
相似文献
1
molecular drug design benchmarking.分子药物设计基准测试
RSC Med Chem. 2021 Jun 3;12(8):1273-1280. doi: 10.1039/d1md00074h. eCollection 2021 Aug 18.
2
GuacaMol: Benchmarking Models for de Novo Molecular Design.GuacaMol：从头设计分子的模型基准测试。
J Chem Inf Model. 2019 Mar 25;59(3):1096-1108. doi: 10.1021/acs.jcim.8b00839. Epub 2019 Mar 19.
3
Fréchet ChemNet Distance: A Metric for Generative Models for Molecules in Drug Discovery.Fréchet ChemNet 距离：药物发现中分子生成模型的一种度量。
J Chem Inf Model. 2018 Sep 24;58(9):1736-1741. doi: 10.1021/acs.jcim.8b00234. Epub 2018 Aug 28.
4
Masked graph modeling for molecule generation.掩蔽图建模用于分子生成。
Nat Commun. 2021 May 26;12(1):3156. doi: 10.1038/s41467-021-23415-2.
5
MolScore: a scoring, evaluation and benchmarking framework for generative models in de novo drug design.MolScore：一种用于从头药物设计中生成模型的评分、评估和基准测试框架。
J Cheminform. 2024 May 30;16(1):64. doi: 10.1186/s13321-024-00861-w.
6
Exploring Tunable Hyperparameters for Deep Neural Networks with Industrial ADME Data Sets.探索具有工业 ADME 数据集的深度神经网络的可调超参数。
J Chem Inf Model. 2019 Mar 25;59(3):1005-1016. doi: 10.1021/acs.jcim.8b00671. Epub 2019 Jan 11.
7
De Novo Peptide and Protein Design Using Generative Adversarial Networks: An Update.使用生成对抗网络的从头肽和蛋白质设计：最新进展
J Chem Inf Model. 2022 Feb 28;62(4):761-774. doi: 10.1021/acs.jcim.1c01361. Epub 2022 Feb 7.
8
Recent Progress of Deep Learning in Drug Discovery.深度学习在药物发现中的最新进展。
Curr Pharm Des. 2021;27(17):2088-2096. doi: 10.2174/1381612827666210129123231.
9
Deep Learning Applied to Ligand-Based De Novo Drug Design.深度学习在配体的从头药物设计中的应用。
Methods Mol Biol. 2022;2390:273-299. doi: 10.1007/978-1-0716-1787-8_12.
10
Molecular Sets (MOSES): A Benchmarking Platform for Molecular Generation Models.分子集（MOSES）：分子生成模型的基准测试平台。
Front Pharmacol. 2020 Dec 18;11:565644. doi: 10.3389/fphar.2020.565644. eCollection 2020.
引用本文的文献
1
AC-ModNet: Molecular Reverse Design Network Based on Attribute Classification.AC-ModNet：基于属性分类的分子反向设计网络。
Int J Mol Sci. 2024 Jun 25;25(13):6940. doi: 10.3390/ijms25136940.
2
TransGEM: a molecule generation model based on Transformer with gene expression data.TransGEM：基于基因表达数据的 Transformer 分子生成模型。
Bioinformatics. 2024 May 2;40(5). doi: 10.1093/bioinformatics/btae189.
3
Automation of Drug Discovery through Cutting-edge Research in Pharmaceuticals: Challenges and Future Scope.通过制药领域的前沿研究实现药物发现自动化：挑战与未来展望。
Curr Comput Aided Drug Des. 2024;20(6):723-735. doi: 10.2174/0115734099260187230921073932.
4
Improving drug discovery with a hybrid deep generative model using reinforcement learning trained on a Bayesian docking approximation.使用基于贝叶斯对接近似的强化学习训练的混合深度生成模型来改进药物发现。
J Comput Aided Mol Des. 2023 Nov;37(11):507-517. doi: 10.1007/s10822-023-00523-3. Epub 2023 Aug 8.
5
To Affinity and Beyond: A Personal Reflection on the Design and Discovery of Drugs.从亲和力到超越：药物设计与发现的个人反思。
Molecules. 2022 Nov 7;27(21):7624. doi: 10.3390/molecules27217624.
6
Comprehensive Survey of Recent Drug Discovery Using Deep Learning.深度学习在药物发现中的最新应用综述
Int J Mol Sci. 2021 Sep 15;22(18):9983. doi: 10.3390/ijms22189983.
本文引用的文献
1
Advances in de Novo Drug Design: From Conventional to Machine Learning Methods.从头药物设计的进展：从传统方法到机器学习方法。
Int J Mol Sci. 2021 Feb 7;22(4):1676. doi: 10.3390/ijms22041676.
2
EvoMol: a flexible and interpretable evolutionary algorithm for unbiased de novo molecular generation.EvoMol：一种用于无偏差从头分子生成的灵活且可解释的进化算法。
J Cheminform. 2020 Sep 16;12(1):55. doi: 10.1186/s13321-020-00458-z.
3
Efficient learning of non-autoregressive graph variational autoencoders for molecular graph generation.用于分子图生成的非自回归图变分自编码器的高效学习。
J Cheminform. 2019 Nov 21;11(1):70. doi: 10.1186/s13321-019-0396-x.
4
Molecular Sets (MOSES): A Benchmarking Platform for Molecular Generation Models.分子集（MOSES）：分子生成模型的基准测试平台。
Front Pharmacol. 2020 Dec 18;11:565644. doi: 10.3389/fphar.2020.565644. eCollection 2020.
5
On failure modes in molecule generation and optimization.关于分子生成和优化中的失效模式。
Drug Discov Today Technol. 2019 Dec;32-33:55-63. doi: 10.1016/j.ddtec.2020.09.003. Epub 2020 Oct 24.
6
OptiMol: Optimization of Binding Affinities in Chemical Space for Drug Discovery.OptiMol：化学空间中用于药物发现的结合亲和力优化。
J Chem Inf Model. 2020 Dec 28;60(12):5658-5666. doi: 10.1021/acs.jcim.0c00833. Epub 2020 Oct 7.
7
A Turing Test for Molecular Generators.分子生成器的图灵测试。
J Med Chem. 2020 Oct 22;63(20):11964-11971. doi: 10.1021/acs.jmedchem.0c01148. Epub 2020 Oct 12.
8
The Synthesizability of Molecules Proposed by Generative Models.生成式模型提出的分子可合成性。
J Chem Inf Model. 2020 Dec 28;60(12):5714-5723. doi: 10.1021/acs.jcim.0c00174. Epub 2020 Apr 17.
9
Bidirectional Molecule Generation with Recurrent Neural Networks.双向分子生成的递归神经网络。
J Chem Inf Model. 2020 Mar 23;60(3):1175-1183. doi: 10.1021/acs.jcim.9b00943. Epub 2020 Jan 16.
10
Efficient multi-objective molecular optimization in a continuous latent space.连续潜在空间中的高效多目标分子优化。
Chem Sci. 2019 Jul 8;10(34):8016-8024. doi: 10.1039/c9sc01928f. eCollection 2019 Sep 14.
Zotero 插件