Sc2Mol: a scaffold-based two-step molecule generator with variational autoencoder and transformer
Open accessAbstract
Motivation: Finding molecules with desired pharmaceutical properties is crucial in drug discovery. Generative models can be an efficient tool to find desired molecules through the distribution learned by the model to approximate given training data. Existing generative models (i) do not consider backbone structures (scaffolds), resulting in inefficiency or (ii) need prior patterns for scaffolds, causing bias. Scaffolds are reasonable to use, and it is imperative to design a generative model without any prior scaffold patterns. Results: We propose a generative model-based molecule generator, Sc2Mol, without any prior scaffold patterns. Sc2Mol uses SMILES strings for molecules. It consists of two steps: scaffold generation and scaffold decoration, which are carried out by a variational autoencoder and a transformer, respectively. The two steps are powerful for implementing random molecule generation and scaffold optimization. Our empirical evaluation using drug-like molecule datasets confirmed the success of our model in distribution learning and molecule optimization. Also, our model could automatically learn the rules to transform coarse scaffolds into sophisticated drug candidates. These rules were consistent with those for current lead optimization. Availability and implementation: The code is available at https://github.com/zhiruiliao/Sc2Mol. Supplementary information: Supplementary data are available at Bioinformatics online.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
Several research studies propose intricate and expansive architectures to overcome the limitations of conditional VAEs.
“To overcome these limitations, several research studies35,36 propose intricate and expansive architectures.”
From PCF-VAE: posterior collapse free variational autoencoder for de novo drug design (2025) · cites this paper as [35]
The Introduction and §2.1 describe Sc2Mol as a two-step architecture combining a VAE with a transformer, so it is an example of a more elaborate molecule generator. Our copy does not say it was proposed specifically to overcome limitations of conditional VAEs; that motivation is unsupported. (LLM, conf=0.82)
Experimental — accuracy not yet measuredThese intricate and expansive architectures attain a SOTA level of validity and diversity in generated molecules but require extended time for completion.
“Although these architectures attain a SOTA level of validity and diversity in generated molecules, the intricate nature of the process leads to an extension in the time required for completion.”
From PCF-VAE: posterior collapse free variational autoencoder for de novo drug design (2025) · cites this paper as [35]
Section 3.5.1, Table 2 reports 100% validity and 99.99% uniqueness for Sc2Mol on MOSES; Section 3.4 identifies uniqueness as a measure of molecular diversity. The cited results support high validity and diversity, but the passages do not establish that architectural complexity extends the time required for completion. (LLM, conf=0.83)
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
(2014) · Cited by 33,419 · see citing works
Advances in Neural Information Processing Systems 29
Teymur, O., Zygalakis, K., & Calderhead, B. (2016) · Cited by 13,881 · see citing works
Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling
Chung, J., Gulcehre, C., Cho, K., & Bengio, Y. (2014) · arXiv · doi:10.48550/arxiv.1412.3555 · Cited by 11,008 · see citing works
Extended-Connectivity Fingerprints
Rogers, D. O. W., & Hahn, M. (2010) · Journal of Chemical Information and Modeling · doi:10.1021/ci100050t · Cited by 7,597 · see citing works
Weininger, D. (1988) · Journal of Chemical Information and Computer Sciences · doi:10.1021/ci00057a005 · Cited by 7,436 · see citing works
A Learning Algorithm for Continually Running Fully Recurrent Neural Networks
Williams, R. J., & Zipser, D. (1989) · Neural Computation · doi:10.1162/neco.1989.1.2.270 · Cited by 4,369 · see citing works
Baell, J. B., & Holloway, G. A. (2010) · Journal of Medicinal Chemistry · doi:10.1021/jm901137j · Cited by 3,963 · see citing works
ZINC 15 – Ligand Discovery for Everyone
Sterling, T., & Irwin, J. J. (2015) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.5b00559 · Cited by 3,224 · see citing works
Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules
Gómez-Bombarelli, R., Wei, J. N., Duvenaud, D., Hernández-Lobato, J. M., Sánchez-Lengeling, B., Sheberla, D., Aguilera-Iparraguirre, J., Hirzel, T. D., Adams, R. P., & Aspuru-Guzik, A. (2018) · ACS Central Science · doi:10.1021/acscentsci.7b00572 · Cited by 3,106 · see citing works
Quantifying the chemical beauty of drugs
Bickerton, G. R., Paolini, G. V., Besnard, J., Mureşan, S., & Hopkins, A. L. (2012) · Nature Chemistry · doi:10.1038/nchem.1243 · Cited by 2,301 · see citing works
Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules.
Gómez‐Bombarelli, R., Wei, J. N., Duvenaud, D., Hernández-Lobato, J. M., Sánchez-Lengeling, B., Sheberla, D., Aguilera‐Iparraguirre, J., Hirzel, T. D., Adams, R. P., & Aspuru‐Guzik, A. (2018) · Cambridge University Engineering Department Publications Database · doi:10.17863/cam.34134 · Cited by 1,813 · see citing works
Prediction of Physicochemical Parameters by Atomic Contributions
Wildman, S. A., & Crippen, G. M. (1999) · Journal of Chemical Information and Computer Sciences · doi:10.1021/ci990307l · Cited by 1,491 · see citing works
Molecular Sets (MOSES): A Benchmarking Platform for Molecular Generation Models
Polykovskiy, D., Zhebrak, A., Sanchez-Lengeling, B., Golovanov, S., Tatanov, O., Belyaev, S., Kurbanov, R., Artamonov, A., Aladinskiy, V., Veselov, M., et al. (2020) · Frontiers in Pharmacology · doi:10.3389/fphar.2020.565644 · Cited by 849 · see citing works
Estimation of the size of drug-like chemical space based on GDB-17 data
Polishchuk, P. G., Madzhidov, T. I., & Varnek, A. (2013) · Journal of Computer-Aided Molecular Design · doi:10.1007/s10822-013-9672-4 · Cited by 619 · see citing works
REINVENT 2.0: An AI Tool for De Novo Drug Design
Blaschke, T., Arús-Pous, J., Chen, H., Margreitter, C., Tyrchan, C., Engkvist, O., Papadopoulos, K., & Patronov, A. (2020) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.0c00915 · Cited by 496 · see citing works
The Necessary Nitrogen Atom: A Versatile High-Impact Design Element for Multiparameter Optimization
Pennington, L. D., & Moustakas, D. T. (2017) · Journal of Medicinal Chemistry · doi:10.1021/acs.jmedchem.6b01807 · Cited by 422 · see citing works
SMILES-based deep generative scaffold decorator for de-novo drug design
Arús‐Pous, J., Patronov, A., Bjerrum, E. J., Tyrchan, C., Reymond, J., Chen, H., & Engkvist, O. (2020) · Journal of Cheminformatics · doi:10.1186/s13321-020-00441-8 · Cited by 204 · see citing works
Where Do Recent Small Molecule Clinical Development Candidates Come From?
Brown, D. G., & Boström, J. (2018) · Journal of Medicinal Chemistry · doi:10.1021/acs.jmedchem.8b00675 · Cited by 179 · see citing works
Scaffold-based molecular design with a graph generative model
Lim, J., Hwang, S.-Y., Moon, S., Kim, S., & Kim, W. Y. (2019) · Chemical Science · doi:10.1039/c9sc04503a · Cited by 172 · see citing works
On failure modes in molecule generation and optimization
Renz, P., Rompaey, D. V., Wegner, J. K., Hochreiter, S., & Klambauer, G. (2019) · Drug Discovery Today Technologies · doi:10.1016/j.ddtec.2020.09.003 · Cited by 154 · see citing works
SyntaLinker: automatic fragment linking with deep conditional transformer neural networks
Yang, Y., Zheng, S., Su, S., Zhao, C., Xu, J., & Chen, H. (2020) · Chemical Science · doi:10.1039/d0sc03126g · Cited by 149 · see citing works
DeepScaffold: A Comprehensive Tool for Scaffold-Based De Novo Drug Discovery Using Deep Learning
Li, Y., Hu, J., Wang, Y., Zhou, J., Zhang, L., & Liu, Z. (2019) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.9b00727 · Cited by 146 · see citing works
Mapping the Efficiency and Physicochemical Trajectories of Successful Optimizations
Young, R. J., & Leeson, P. D. (2018) · Journal of Medicinal Chemistry · doi:10.1021/acs.jmedchem.8b00180 · Cited by 112 · see citing works
Scaffold-Constrained Molecular Generation
Langevin, M., Minoux, H., Levesque, M., & Bianciotto, M. (2020) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.0c01015 · Cited by 67 · see citing works
Felts, A. S., Rodriguez, A. L., Blobaum, A. L., Morrison, R. D., Bates, B. S., Gray, A. T., Rook, J. M., Tantawy, M. N., Byers, F. W., Chang, S., et al. (2017) · Journal of Medicinal Chemistry · doi:10.1021/acs.jmedchem.7b00410 · Cited by 34 · see citing works
Felts, A. S., Rodriguez, A. L., Morrison, R. D., Venable, D. F., Manka, J. T., Bates, B. S., Blobaum, A. L., Byers, F. W., Daniels, J. S., Niswender, C. M., et al. (2013) · Bioorganic & Medicinal Chemistry Letters · doi:10.1016/j.bmcl.2013.09.001 · Cited by 33 · see citing works
Zhang, K. Y. J., Milburn, M. V., & Artis, D. R. (2007) · Springer eBooks · doi:10.1007/1-4020-4407-0_6 · Cited by 11 · see citing works
Bates, B. S., Rodriguez, A. L., Felts, A. S., Morrison, R. D., Venable, D. F., Blobaum, A. L., Byers, F. W., Lawson, K. P., Daniels, J. S., Niswender, C. M., et al. (2014) · Bioorganic & Medicinal Chemistry Letters · doi:10.1016/j.bmcl.2014.06.003 · Cited by 9 · see citing works
Flagged first, then by citations
Research works printed but not linked: 13
- Arjovsky M. et al (2017) Wasserstein Generative Adversarial Networks. In: Proceedings of the 34th International Conference on Machine Learning, Sydney, Australia, pp. 214– 223.
- Bowman S.R. et al (2016) Generating sentences from a continuous space. In: Proceedings of the 20th SIGNLL Conference on Computational Natural Language Learning, pp. 10– 21. Association for Computational Linguistics, Berlin, Germany.
- Dauphin Y.N. et al (2017) Language modeling with gated convolutional networks. In: Proceedings of the 34th International Conference on Machine Learning, Sydney, Australia, pp. 933– 941.
- Dong H.-W. et al (2018) MuseGAN: multi-track sequential generative adversarial networks for symbolic music generation and accompaniment. In: Proceedings of the AAAI Conference on Artificial Intelligence, New Orleans, Louisiana, USA, Vol. 32.
- He K. et al (2016) Deep residual learning for image recognition. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), pp. 770– 778. IEEE Computer Society, Los Alamitos, CA, USA.
- Jin W. et al (2018) Junction tree variational autoencoder for molecular graph generation. In: Dy J. and Krause A., editors, Proceedings of the 35th International Conference on Machine Learning, Volume 80 of Proceedings of Machine Learning Research. PMLR, pp. 2323– 2332.
- Karras T. et al (2020) Analyzing and improving the image quality of StyleGAN. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, pp. 8110-8119.
- Kingma D.P., Ba J. (2015) Adam: a method for stochastic optimization. In: Proceedings of3rd International Conference on Learning Representations, San Diego, CA, USA.
- Kingma D.P., Welling M. (2014) Auto-encoding variational Bayes. In: Proceedings of2nd International Conference on Learning Representations, Banff, AB, Canada.
- Lei Ba J. et al (2016) Layer normalization. arXiv preprint arXiv:1607.06450. · arXiv
- Zang C., Wang F. (2020) Moflow: an invertible flow model for generating molecular graphs. In: Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, New York, NY, USA, pp. 617– 626.
- Zhang H. et al (2019) ReCoSa: detecting the relevant contexts with self-attention for multi-turn dialogue generation. In: Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics, pp. 3721– 3730. Association for Computational Linguistics, Florence, Italy
- Zhang J. et al (2017) Flexible and creative Chinese poetry generation using neural memory. In: Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, Volume 1: Long Papers), pp. 1364– 1373. Association for Computational Linguistics, Vancouver, Canada. PMC12007664 40255468
Cited by
Sorted by citations
Unleashing the power of generative AI in drug discovery
Gangwal, A., & Lavecchia, A. (2024) · Drug Discovery Today · doi:10.1016/j.drudis.2024.103992 · Cited by 195 · see citing works
Gotta be SAFE: a new framework for molecular design
Noutahi, E., Gabellini, C., Craig, M. J., Lim, J. S. C., & Tossou, P. (2024) · Digital Discovery · doi:10.1039/d4dd00019f · Cited by 52 · see citing works
A systematic review of deep learning chemical language models in recent era
Flores-Hernandez, H., & Martínez-Ledesma, E. (2024) · Journal of Cheminformatics · doi:10.1186/s13321-024-00916-y · Cited by 46 · see citing works
Liu, C.-S., Yan, B., Sun, H., Lu, J., & Puno, P. (2025) · Natural Products and Bioprospecting · doi:10.1007/s13659-025-00521-y · Cited by 23 · see citing works
MVML-MPI: Multi-View Multi-Label Learning for Metabolic Pathway Inference
Liu, X., Yang, H., Ai, C., Ding, Y., Guo, F., & Tang, J. (2023) · Briefings in Bioinformatics · doi:10.1093/bib/bbad393 · Cited by 16 · see citing works
Improving the Reliability of Molecular String Representations for Generative Chemistry
Reboul, E., Wefers, Z., Prabakaran, H., Waldispühl, J., & Taly, A. (2025) · Journal of Chemical Information and Modeling · doi:10.1021/acs.jcim.4c02261 · Cited by 13 · see citing works
Machine Learning for De Novo Molecular Generation: A Comprehensive Review
Chen, Y., & Xue, W. (2026) · ACS Chemical Neuroscience · doi:10.1021/acschemneuro.5c00861 · Cited by 12 · see citing works
Application progress of deep generative models in de novo drug design
Liu, Y., Xu, C., Yang, X., Zhang, Y., Chen, Y., & Liu, H. (2024) · Molecular Diversity · doi:10.1007/s11030-024-10942-5 · Cited by 11 · see citing works
Bhadwal, A. S., Kumar, K., & Kumar, N. (2023) · Expert Systems with Applications · doi:10.1016/j.eswa.2023.122396 · Cited by 13 · see citing works
PCF-VAE: posterior collapse free variational autoencoder for de novo drug design
Bhadwal, A. S., Kumari, M., & Kumar, A. (2025) · Scientific Reports · doi:10.1038/s41598-025-14285-5 · Cited by 7 · see citing works
LOGICS: Learning optimal generative distribution for designing de novo chemical structures
Bae, B., Bae, H., & Nam, H. (2023) · Journal of Cheminformatics · doi:10.1186/s13321-023-00747-3 · Cited by 7 · see citing works
Liu, D., Liu, D., Sheng, K., Cheng, Z., Liu, Z., Qiao, Y., Cai, S., Li, Y., Wang, J., Chen, H., et al. (2024) · Briefings in Bioinformatics · doi:10.1093/bib/bbae690 · Cited by 6 · see citing works
Exploring Chemical Space with Chemistry-Inspired Dynamic Quantum Circuits in the NISQ Era
Chen, L.-Y., Li, T.-Y., Li, Y., Chen, N.-Y., & You, F. (2025) · Journal of Chemical Theory and Computation · doi:10.1021/acs.jctc.5c00305 · Cited by 5 · see citing works
A comprehensive survey on artificial intelligence for biomolecule design
Yang, G., Li, J., Wang, Q., Wu, L., Chen, Z., Zhao, P.-C., Wang, H., Wang, Q., Wen, B., Song, C., et al. (2026) · The Innovation Life · doi:10.59717/j.xinn-life.2026.100196 · Cited by 5 · see citing works
Dagher, J., Nassar, M., & Faour, W. H. (2025) · European Journal of Pharmacology · doi:10.1016/j.ejphar.2025.178183 · Cited by 4 · see citing works
De novo Molecular Design using Deep Learning
Nguyen‐Vo, T. (2024) · doi:10.26686/wgtn.25044362 · Cited by 4 · see citing works
Enhancing Unconditional Molecule Generation via Online Knowledge Distillation of Scaffolds
Wang, H., Wang, Z., Shi, M., Cheng, Z., & Qian, Y. (2025) · Molecules · doi:10.3390/molecules30061262 · Cited by 3 · see citing works
Scaffold-driven molecular generation via reinforced RNN with centroid distance evaluation
Zhu, X., Zhao, Z., & Zhu, F. (2025) · Expert Systems with Applications · doi:10.1016/j.eswa.2025.128606
Zhou, Y., Sang, Z., Xu, C., Cao, Z., Xiao, K., Jia, Q., He, Y., Luo, H., & Huang, S. (2025) · Molecular Diversity · doi:10.1007/s11030-025-11342-z
Han, M., Li, B., Li, X., Zhang, H., & Li, B. (2025) · International Conference on Bioinformatics and Biomedical Technology (ICBBT) · doi:10.1109/icbbt65815.2025.11276459
Bae, B., Bae, H., & Nam, H. (2024) · Figshare · doi:10.6084/m9.figshare.26615515
Zhang, Y., Fu, Y., & Wang, G. (2026) · Journal of Computer-Aided Molecular Design · doi:10.1007/s10822-026-00887-2
CCVAN Leverages Conditional Molecular Generation Through Conditional VAE and Wasserstein GAN
Zheng, J., Xu, Z., Huang, J., Chen, X., Li, Y., Wei, Y., & Zhang, H. (2026) · Interdisciplinary Sciences Computational Life Sciences · doi:10.1007/s12539-026-00860-5
Scaling Biomedical Text-Mining: Transformers, GenAI, and Drug Discovery
Agostinho, S. P., Santos, C., Ramalhão, D., Moreira, I. S., & Rosário‐Ferreira, N. (2026) · Methods in molecular biology · doi:10.1007/978-1-0716-5408-8_4
- Authors
- Liao, Z., Xie, L., Mamitsuka, H., & Zhu, S.
- Affiliations
- Fudan University (CN)
- City University of New York (US)
- Kyoto University (JP)
- Corresponding author
- Zhu, Shanfeng
- Venue
- Bioinformatics
- Publisher
- Oxford University Press (OUP)
- Published
- 2022-12-28
- Volume
- 39
- Issue
- 1
- ISSN
- 1367-4803 (ISSN-L)
- 1367-4811
- PMC
- PMC9835482
- OpenAlex
- W4313236609
- Cited by
- 24
- References
- 28
- Citations by year
- 2026: 5 · 2025: 9 · 2024: 7 · 2023: 3
- FWCI
- 2.57
- Type
- journal article
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- Locations
- Bioinformatics (publishedVersion, cc-by)
- PubMed Central (publishedVersion, cc-by)
- PubMed (publishedVersion)
- Institutional Repositories DataBase (IRDB) (submittedVersion, cc-by)
- Aaltodoc (Aalto University) (submittedVersion, cc-by)
- Kyoto University Research Information Repository (Kyoto University) (submittedVersion, cc-by)
- Funders
- National Natural Science Foundation of China (62272105)
- Topics
- Physical Sciences › Computer Science › Computational Theory and Mathematics › Computational Drug Discovery Methods
- Physical Sciences › Materials Science › Materials Chemistry › Machine Learning in Materials Science
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Protein Degradation and Inhibitors
- Keywords
- molecular generation, scaffold optimization, variational autoencoder, transformer models, SMILES, drug-like molecules, lead optimization
- MeSH terms
- Machine Learning, Drug Discovery
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
