Academic Journal
Προβολή σε EDS

Property-Conditioned Transformer Language Model for De Novo Molecular Generation Using Self-Referencing Embedded Strings.

Λεπτομέρειες βιβλιογραφικής εγγραφής
Τίτλος: Property-Conditioned Transformer Language Model for De Novo Molecular Generation Using Self-Referencing Embedded Strings.
Συγγραφείς: Patel A; Department of Pharmaceutical Chemistry, Parul Institute of Pharmacy, Parul University, Vadodara, Gujarat, India., Maheshwari RA; Department of Pharmacy, Sumandeep Vidyapeeth Deemed to be University, Vadodara, Gujarat, India., Patel A; Charotar University of Science and Technology, CHARUSAT-Campus, Anand, Gujarat, India., Faldu A; Smt. Chandaben Mohanbhai Patel Institute of Computer Applications, Charotar University of Science and Technology, Anand, Gujarat, India., Mahavar A; Charotar University of Science and Technology, CHARUSAT-Campus, Anand, Gujarat, India.
Πηγή: Molecular informatics [Mol Inform] 2026 Oct; Vol. 45 (10), pp. e70055.
Τύπος έκδοσης: Journal Article
Γλώσσα: English
Στοιχεία περιοδικού: Publisher: Wiley-VCH Verlag Country of Publication: Germany NLM ID: 101529315 Publication Model: Print Cited Medium: Internet ISSN: 1868-1751 (Electronic) Linking ISSN: 18681743 NLM ISO Abbreviation: Mol Inform Subsets: MEDLINE
Imprint Name(s): Original Publication: Weinheim, Germany : Wiley-VCH Verlag
Ιατρικοί όροι (MeSH): Models, Molecular*, Multilayer Perceptrons ; Reinforcement Machine Learning
Περίληψη: Transformer-based molecular language models have demonstrated strong capabilities for de novo molecular generation by modeling chemical structures as token sequences. However, controllable generation of molecules with desired physicochemical properties remains challenging without reinforcement learning (RL) or post-generation filtering. In this study, we propose a property-conditioned autoregressive transformer framework for molecular generation using the SELFIES representation. Continuous molecular descriptors, including logP, quantitative estimate of drug-likeness, and molecular weight (MolWt), were normalized and embedded via a multilayer perceptron and injected into token embeddings to enable global conditioning during sequence modeling. The model was trained end-to-end using the maximum likelihood with cross-entropy loss. Experiments conducted on the combined subsets of QM9, ZINC, and MOSES datasets demonstrated high validity (100%), strong uniqueness, and substantial novelty in the generated molecules. Furthermore, the generated property distributions closely aligned with the training data statistics. The proposed framework provides a computationally efficient and stable approach for controllable molecular generation without requiring RL or graph-based modeling.
(© 2026 Wiley‐VCH GmbH.)
References: G. Schneider, “Automating Drug Discovery,” Nature Reviews Drug Discovery 17, no. 2 (2018): 97–113, https://doi.org/10.1038/nrd.2017.232.
N. Brown, P. Ertl, R. Lewis, T. Luksch, D. Reker, and N. Schneider, “Artificial Intelligence in Chemistry and Drug Design,” Journal of Computer‐Aided Molecular Design 34, no. 7 (2020): 709–715, https://doi.org/10.1007/s10822‐020‐00317‐x.
J. Vamathevan, D. Clark, P. Czodrowski, et al., “Applications of Machine Learning in Drug Discovery and Development,” Nature Reviews Drug Discovery 18, no. 6 (2019): 463–477, https://doi.org/10.1038/s41573‐019‐0024‐5.
A. Zhavoronkov, Y. A. Ivanenkov, A. Aliper, et al., “Deep Learning Enables Rapid Identification of Potent DDR1 Kinase Inhibitors,” Nature Biotechnology 37, no. 9 (2019): 1038–1040, https://doi.org/10.1038/s41587‐019‐0224‐x.
B. Sanchez‐Lengeling and A. Aspuru‐Guzik, “Inverse Molecular Design Using Machine Learning: Generative Models for Matter Engineering,” Science 361, no. 6400 (2018): 360–365, https://doi.org/10.1126/science.aat2663.
W. Gao and C. W. Coley, The Synthesizability of Molecules Proposed by Generative Models, 2020, arXiv [q‐Bio.QM], http://arxiv.org/abs/2002.07007.
W. P. Walters and M. Murcko, “Assessing the Impact of Generative AI on Medicinal Chemistry,” Nature Biotechnology 38, no. 2 (2020): 143–145, https://doi.org/10.1038/s41587‐020‐0418‐2.
A. Djennad, R. Kherbich, I. Youkana, L. Kahloul, and R. Saouli, Advances in Intelligent Systems Research, (Atlantis Press International, 2025), 56–64, https://doi.org/10.2991/978‐94‐6463‐805‐9_8.
C. Cerchia and A. Lavecchia, “New Avenues in Artificial‐Intelligence‐Assisted Drug Discovery,” Drug Discovery Today 28, no. 4 (2023): 103516, https://doi.org/10.1016/j.drudis.2023.103516.
T. Kumar, and P. Kavya S, “AI‐Driven Molecular Design and Screening for Drug Discovery,” International Journal of Scientific Research in Engineering and Management 9, no. 8 (2025): 1–9, https://doi.org/10.55041/ijsrem52104.
U. Das, “Generative AI for Drug Discovery and Protein Design: The Next Frontier in AI‐Driven Molecular Science,” Medicine in Drug Discovery 27 (2025): 100213, https://doi.org/10.1016/j.medidd.2025.100213.
H. Abbad, M. Taieb Brahim, L. Oulladji, and L. Abbad, “AI‐Driven Molecular Design: Synergizing Deep Generative Models With Evolutionary Optimization. The Eurasia Proceedings of Science, Technology,” Engineering and Mathematics 38 (2026): 512–524, https://doi.org/10.55549/epstem.1248.
K. T. Butler, D. W. Davies, H. Cartwright, O. Isayev, and A. Walsh, “Machine Learning for Molecular and Materials Science,” Nature 559 (2018): 547–555, https://www.nature.com/articles/s41586‐018‐0337‐2.
A. Tkatchenko, “Machine Learning for Chemical Discovery,” Nature Communications 11 (2023): 1–4, https://www.nature.com/articles/s41467‐020‐17844‐8.
Y. Chen, X. Guo, and W. Xue, “Large Language Models for Molecular Design: Bridging the Gap Between Chemical Syntax and Biological Semantics,” Drug Discovery Today 31, no. 2 (2026): 104634, https://doi.org/10.1016/j.drudis.2026.104634.
X.‐D. Wang, Z.‐R. Chen, W.‐K. Li, et al., “Artificial Intelligence for Perovskite Additive Engineering: From Molecular Screening to Autonomous Discovery,” Molecules 31, no. 3 (2026): 440, https://doi.org/10.3390/molecules31030440.
D. C. Elton, Z. Boukouvalas, M. D. Fuge, and P. W. Chung, “Deep Learning for Molecular Design—a Review of the State of the Art,” Molecular Systems Design & Engineering 4, no. 4 (2019): 828–849, https://doi.org/10.1039/c9me00039a.
C. Zhao, J. Lin, Y. Liu, et al., “Generative Deep Learning‐Aided Design of Flexible Molecular Crystals,” Chemistry of Materials 37, no. 20 (2025): 8421–8435, https://doi.org/10.1021/acs.chemmater.5c02194.
W. Hussain, N. Rasool, and Y. D. Khan, “Insights Into Machine Learning‐Based Approaches for Virtual Screening in Drug Discovery: Existing Strategies and Streamlining Through FP‐CADD,” Current Drug Discovery Technologies 18, no. 4 (2021): 463–472, https://doi.org/10.2174/1570163817666200806165934.
J. Mao, J. Wang, A. Zeb, et al., “Transformer‐Based Molecular Generative Model for Antiviral Drug Design,” Journal of Chemical Information and Modeling 64, no. 7 (2024): 2733–2745, https://doi.org/10.1021/acs.jcim.3c00536.
J. Wang, “Autoregressive Language Models for Molecular Generation,” Journal of Chemical Information and Modeling 64 (2024): 2125–2136.
R. Abe, and T. Miyao, “Long Short‐Term Memory‐Based Chemical Language Models for Bioactive Molecular Generation Using Tailored Pre‐Training Datasets,” Artificial Intelligence in the Life Sciences 9 (2026): 100159, https://doi.org/10.1016/j.ailsci.2026.100159.
A. Yüksel, E. Ulusoy, A. Ünlü, and T. Doğan, SELFormer: Molecular Representation Learning via SELFIES Language Models, 2023, arXiv [q‐Bio.QM], https://doi.org/10.48550/arXiv.2304.04662.
D. Huang, and S. Tu, “MulMol: Transformer‐based Multi‐Task Molecular Representation Learning,” in 2024 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), (IEEE, 2024), 681–686, https://doi.org/10.1109/BIBM62325.2024.10822716.
Q. Hu, C. Sun, H. He, et al., “Target‐Aware 3D Molecular Generation Based on Guided Equivariant Diffusion,” Nature Communications 16, no. 1 (2025): 7928, https://doi.org/10.1038/s41467‐025‐63245‐0.
J. Jiang, L. Ke, L. Chen, et al., “Transformer Technology in Molecular Science,” WIREs Computational Molecular Science 14, no. 4 (2024): e1725, https://doi.org/10.1002/wcms.1725.
Y. Wu, Y. Yang, R. Zhang, et al., “Molecular Representation Learning Based on Transformer With Fixed‐Length Padding Method,” Journal of Molecular Structure 1319 (2025): 139574, https://doi.org/10.1016/j.molstruc.2024.139574.
A. Sultan, M. Rausch‐Dupont, S. Khan, O. Kalinina, D. Klakow, and A. Volkamer, A. Transformers for Molecular Property Prediction: Domain Adaptation Efficiently Improves Performance, 2025, arXiv [cs.LG], https://doi.org/10.48550/arXiv.2503.03360.
V. Bagal, R. Aggarwal, P. K. Vinod, and U. D. Priyakumar, “MolGPT: Molecular Generation Using a Transformer‐Decoder Model,” Journal of Chemical Information and Modeling 62, no. 9 (2022): 2064–2076, https://doi.org/10.1021/acs.jcim.1c00600.
C. Abate, S. Decherchi, and A. Cavalli, “Graph Neural Networks for Conditional De Novo Drug Design,” WIREs Computational Molecular Science 13 (2023): e1651, https://doi.org/10.1002/wcms.1651.
H. Ma, Y. Bian, Y. Rong, et al., Multi‐View Graph Neural Networks for Molecular Property Prediction, 2020, arXiv [q‐Bio.QM], http://arxiv.org/abs/2005.13607.
Y. Zhao, Y. Chen, J. Du, et al., “Dual‐Route Embedding‐Aware Graph Neural Networks for Drug Repositioning,” Briefings in Bioinformatics 26, no. 5 (2025): bbaf555, https://doi.org/10.1093/bib/bbaf555.
J. Chen, N. Fan, Y. Lu, et al., “Intelligence on the Graph: Graph Neural Networks for Mechanistic Drug Target Discovery,” Journal of Pharmaceutical Analysis 16, no. 7 (2026): 101508, https://doi.org/10.1016/j.jpha.2025.101508.
K. Berry and L. Cheng, A Survey of Graph Neural Networks for Drug Discovery: Recent Developments and Challenges, 2025, arXiv [cs.LG], http://arxiv.org/abs/2509.07887.
P. Bongini, “Composite Graph Neural Networks for Molecular Modeling,” International Journal of Molecular Sciences 25, 2024): 6583, https://doi.org/10.3390/ijms25126583.
Y. Jing, D. Zhang, and L. Li, “H2GnnDTI: Hierarchical Heterogeneous Graph Neural Networks for Drug‐Target Interaction Prediction,” Bioinformatics 41, no. 4 (2025): btaf117, https://doi.org/10.1093/bioinformatics/btaf117.
X. Lin, L. Dai, Y. Zhou, et al., “Comprehensive Evaluation of Deep and Graph Learning on Drug‐Drug Interactions Prediction,” Briefings in Bioinformatics 24, no. 4 (2023): bbad235, https://doi.org/10.1093/bib/bbad235.
Z. Zhang, J. Yan, Y. Huang, et al., “Structure‐Based Drug Design With Geometric Deep Learning: A Comprehensive Survey,” ACM Computing Surveys 58, no. 5 (2026): 1–35, https://doi.org/10.1145/3769677.
W. Y. Kim, H. Kim, S. Moon, W. Zhung, and J. Lim, DeepBioisostere: Deep Learning‐Based Bioisosteric Replacements for Optimization of Multiple Molecular Properties, 2024, Research Square, https://doi.org/10.21203/rs.3.rs‐4225249/v1.
A. A. Singh, Comprehensive Benchmarking Platform for Deep Generative Models in Molecular Design, 2025, arXiv [physics.atom‐Ph], http://arxiv.org/abs/2505.12848.
P. Reiser, M. Neubert, A. Eberhard, et al., “Graph Neural Networks for Materials Science and Chemistry,” Communications Materials 3, no. 1 (2022): 93, https://doi.org/10.1038/s43246‐022‐00315‐6.
K. Zhang, X. Wen, J. Xiang, X. Wang, Y. Gao, and R. Cao, “Multi‐Representation Local‐Global Deep Learning Architecture for Molecular Property Prediction,” in 2025 International Joint Conference on Neural Networks (IJCNN), (IEEE, 2025), 1–8, https://doi.org/10.1109/IJCNN64981.2025.11228160.
A. Repetto, G. Ramis, and I. Rossetti, “Machine Learning and Approximated Estimation Approaches for Process Design in Drug Synthesis,“ Chemistry 8, no. 3 (2026): 32, https://doi.org/10.3390/chemistry8030032.
D. Polykovskiy, A. Zhebrak, B. Sanchez‐Lengeling, et al., “Molecular Sets (MOSES): A Benchmarking Platform for Molecular Generation Models,” Frontiers in Pharmacology 11 (2020): 565644, https://doi.org/10.3389/fphar.2020.565644.
T. Sterling and J. J. Irwin, “ZINC 20‐a Free Database of Commercially Available Compounds,” Journal of Chemical Information and Modeling 63 (2023): 3891–3898, https://doi.org/10.1021/acs.jcim.0c00675.
M. Krenn, “SELFIES: A Robust Representation of Semantically Constrained Graphs With an Example Application in Chemistry,” Machine Learning: Science and Technology 1 (2022): 045024, https://arxiv.org/abs/1905.13741v1.
G. Landrum, “RDKit: Open‐Source Cheminformatics Software,” Journal of Chemical Information and Modeling 63 (2023): 3450–3456.
A. Paszke, S. Gross, F. Massa, et al., PyTorch: An Imperative Style, High‐Performance Deep Learning Library, 2019, arXiv [cs.LG], http://arxiv.org/abs/1912.01703.
D. P. Kingma and J. Ba, Adam: A Method for Stochastic Optimization, 2014, arXiv [cs.LG], https://doi.org/10.48550/arXiv.1412.6980.
Y. Du, A. R. Jamasb, J. Guo, et al., “Machine Learning‐Aided Generative Molecular Design,” Nature Machine Intelligence 6, no. 6 (2024): 589–604, https://doi.org/10.1038/s42256‐024‐00843‐5.
H. Yang, Q. Zhong, M. Wang, J. Peng, and D. Shen, “3D Molecule Generation via Diffusion Model With a Self‐Attention‐Based EGNN,” ACS Omega 10 (2025): 57813–57829, https://doi.org/10.1021/acsomega.5c10621.
C. Wang, G. A. Kumar, and J. C. Rajapakse, “Drug Discovery and Mechanism Prediction With Explainable Graph Neural Networks,” Scientific Reports 15, no. 1 (2025): 179, https://doi.org/10.1038/s41598‐024‐83090‐3.
Y. Liu, “Transformer Model for Molecular Generation From Gene Expression,” Bioinformatics 40 (2024): btae189, https://doi.org/10.1093/bioinformatics/btae189.
J. Ding, S. Jin, Y. Meng, Q. Jin, and J. Xu, “TransFVAE: A Transformer‐Based Flow Variational Autoencoder Model for Molecular Graph Generation,“ in 2025 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), (IEEE,2025), 645–651, https://doi.org/10.1109/BIBM66473.2025.11356197.
N. Joshi and E. Guven, Improved Molecular Generation Through Attribute‐Driven Integrative Embeddings and GAN Selectivity, 2025, arXiv [cs.LG], https://doi.org/10.48550/arXiv.2504.19040.
D. Manu, J. Yao, W. Liu, and X. Sun, GraphGANFed: A Federated Generative Framework for Graph‐Structured Molecules Towards Efficient Drug Discovery, 2023, arXiv [cs.LG], https://doi.org/10.48550/arXiv.2304.05498.
P. Gaiński, M. Koziarski, J. Tabor, and M. Śmieja, ChiENN: Embracing Molecular Chirality With Graph Neural Networks, 2023, arXiv [cs.LG], https://doi.org/10.48550/arXiv.2307.02198.
J. Wee and J. Jiang, A Review of Topological Data Analysis and Topological Deep Learning in Molecular Sciences, 2025,arXiv [q‐Bio.BM], http://arxiv.org/abs/2509.16877.
K. Atz, D. F. Nippa, A. T. Müller, et al., “Geometric Deep Learning‐Guided Suzuki Reaction Conditions Assessment for Applications in Medicinal Chemistry,” RSC Medicinal Chemistry 15, no. 7 (2024): 2310–2321, https://doi.org/10.1039/d4md00196f.
Y. Sun, L. Chen, Z. Jing, et al., “Generative AI for the Design of Molecules: Advances and Challenges,” Journal of Chemical Information and Modeling 65, no. 23 (2025): 12668–12690, https://doi.org/10.1021/acs.jcim.5c02234.
Y. Shi, S. Yang, W. Li, Y. Wu, and W. Luo, “Deep Learning‐Driven Intelligent Fluorescent Probes: Advancements in Molecular Design for Accurate Food Safety Detection,” Foods 14, no. 17 (2025): 3114, https://doi.org/10.3390/foods14173114.
H. Wang, Z. Zhang, and H. Gong, qcMol: A Large‐Scale Dataset of 1.2 Million Molecules With High‐Quality Quantum Chemical Annotations for Molecular Representation Learning, 2025, bioRxiv, https://doi.org/10.1101/2025.09.07.674462.
H. Li, S. Sarkar, W. Lu, et al., “Collective Intelligence for AI‐Assisted Chemical Synthesis,” Nature 651, no. 8104 (2026): 107–115, https://doi.org/10.1038/s41586‐026‐10131‐4.
J. Zhou, and M. Chen, “Orthogonal Molecular Feature Signatures Guide Multi‐Target Alzheimer's Drug Discovery Through Graph Transformer Representation Learning,” Journal of Dementia and Alzheimer's Disease 3, no. 2 (2026): 19, https://doi.org/10.3390/jdad3020019.
X. Lu, L. Xie, L. Xu, R. Mao, X. Xu, and S. Chang, “Multimodal Fused Deep Learning for Drug Property Prediction: Integrating Chemical Language and Molecular Graph,” Computational and Structural Biotechnology Journal 23 (2024): 1666–1679, https://doi.org/10.1016/j.csbj.2024.04.030.
Contributed Indexing: Keywords: artificial intelligence; autoregressive model; computer science; graph; language model; molecular descriptor; perceptron; reinforcement learning; training set; transformer
Entry Date(s): Date Created: 20260929 Date Completed: 20260929 Latest Revision: 20260929
Update Code: 20260929
DOI: 10.1002/minf.70055
PMID: 42806511
Βάση Δεδομένων: MEDLINE
Περιγραφή
ISSN:1868-1751
DOI:10.1002/minf.70055