Pith. sign in

REVIEW 5 major objections 5 minor 57 references

A Comprehensive Benchmarking Platform for Deep Generative Models in Molecular Design

T0 review · 5 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read By running six deep generative models through the MOSES benchmark, this paper establishes that no architecture dominates molecular design: CharRNN is best at matching the overall distribution of drug-like molecules while VAE is best at…

desk verdict This is a repackaged MOSES benchmark run, not a new study — the Table I values are the published MOSES numbers, unreferenced and without protocol or error bars. read the letter →

arxiv 2505.12848 v1 pith:MTOAE66G submitted 2025-05-19 physics.atom-ph cs.LG

Adarsh Singh This is my paper · ORCID
classification physics.atom-phcs.LG
keywords moleculargenerationbenchmarkingMOSESdeepgenerativemodelsdrugdiscoveryvariationalautoencoderrecurrentneuralnetworkevaluationmetrics
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Using the MOSES benchmark, this paper compares six deep generative models—CharRNN, VAE, AAE, JTN-VAE, and LatentGAN, plus three classical baselines—on their ability to generate valid, unique, and novel drug-like molecules while preserving chemical properties. The central finding is that no architecture wins on every metric: CharRNN reproduces the overall distribution of generated molecules best, while VAE best preserves molecular scaffolds and property distributions. These complementary strengths are interpreted as an exploration–exploitation trade-off in chemical space, with more novel models such as LatentGAN sacrificing distribution fidelity. The paper argues that model selection should therefore depend on the drug-discovery stage, such as lead optimization versus scaffold hopping.

What carries the argument

The load-bearing object is the MOSES evaluation suite: a filtered database of lead-like molecules split into training, test, and scaffold-separated test sets, paired with a fixed battery of metrics—validity, uniqueness, novelty, FCD, SNN, scaffold similarity, internal diversity, and Wasserstein-1 distances on four physicochemical properties. The mechanism is that each metric isolates one axis of quality, so the joint profile of a model shows where it trades off distribution matching against exploration. To make the comparison, the paper re-implements six generative architectures on identical data and supplements them with property-prediction models that quantify how well generated molecules can be predictably optimized.

What would settle it

Train the same six models on the MOSES training set, generate a fixed number of molecules from each, and evaluate them with a distribution distance computed from 3D pharmacophore fingerprints instead of the SMILES-based FCD. If the rank order of CharRNN and VAE on the new distance differs from their FCD rank order, the claimed exploration–exploitation trade-off is metric-dependent rather than a fixed property of the models.

Watch

Extended reading notes

Core claim

The paper's claim is that the MOSES platform's standardized evaluation exposes a structured trade-off landscape in molecular generation. Across the six models, validity, uniqueness, novelty, FCD, SNN, scaffold similarity, and Wasserstein distances for logP, SA, QED, and molecular weight define a multi-dimensional profile. CharRNN attains the lowest FCD (0.073) and near-perfect uniqueness, VAE attains the highest scaffold similarity (0.939) and SNN (0.626) with the lowest deep-model novelty (69.5%), JTN-VAE guarantees 100% validity while reaching 91.4% novelty, and LatentGAN reaches 95.0% novelty with higher FCD. The paper concludes that these complementary strengths reflect fundamental trade-offs between exploring novel chemical space and staying close to the training distribution, so no single architecture is universally best.

Load-bearing premise

The comparison assumes the MOSES metric suite is a faithful characterization of generative-model quality for drug discovery, so the observed trade-offs are properties of the models rather than artifacts of these particular scores.

Editorial extensions

If this is right

  • For lead-optimization campaigns, VAE-type models are preferable because their high scaffold similarity and low novelty keep generations close to known active compounds.
  • For de novo design and scaffold hopping, LatentGAN and JTN-VAE provide more novel structures, with JTN-VAE adding guaranteed validity at the cost of higher computational expense.
  • Hybrid architectures should be explored, e.g., coupling JTN-VAE scaffold generation with CharRNN refinement, to combine validity guarantees with strong distribution matching.
  • Model evaluation should report the full MOSES profile rather than a single metric, since validity alone obscures trade-offs in novelty and distribution fidelity.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The novelty-versus-fit trade-off is partly metric-built: FCD penalizes distance from the test set while novelty rewards distance from the training set, so their negative correlation may overstate a real chemical-space tension.
  • A concrete test: generate molecules from VAE latent-space interpolations and train a CharRNN on them; if the fine-tuned model keeps FCD low with higher novelty, the trade-off is breakable.
  • Adding a scaffold-novelty score on the MOSES scaffold test set would separate scaffold exploration from whole-molecule exploration, giving a cleaner measure of exploration.
  • Because distribution matching alone says nothing about biological activity, pairing the benchmark with target-specific docking enrichment would show which model's exploration yields useful chemistry.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

5 major / 5 minor

Summary. The paper claims to provide a comprehensive benchmarking evaluation of deep generative models for molecular design using the MOSES platform. It describes the MOSES dataset, several generative model families (CharRNN, VAE, AAE, JTN-VAE, LatentGAN, and classical baselines), a suite of evaluation metrics, and property prediction models. The central empirical claim, stated in the abstract and conclusion, is that different architectures show complementary strengths and that CharRNN and VAE are the strongest overall performers, with CharRNN best at distribution matching and VAE best at property preservation. The paper also discusses trade-offs including validity versus novelty, distribution matching versus diversity, and exploration versus exploitation.

Significance. If the empirical results were trustworthy, the observation of complementary strengths and the emphasis on standardized benchmarking would be a useful synthesis for practitioners choosing generative models. However, the manuscript presents no new methods, no new dataset, and no new code; the reported numbers appear to reproduce the original MOSES benchmark table, yet this is not stated. The main significance would therefore be as a review or reproduction, but the current presentation claims an original comprehensive evaluation without supplying the necessary protocol, uncertainty quantification, or reproducible artifacts. The paper does not provide machine-checked proofs, reproducible code, or parameter-free derivations, so its value rests entirely on the reliability of the reported experimental numbers, which the manuscript does not support.

major comments (5)
  1. [§IV-A, Table I] The central trade-off claim rests on Table I, but the table is garbled: the columns for Novelty, SNN, Scaff, and other metrics are interleaved with the text rather than presented as a clean table, and no evaluation protocol is given. Section III-B and III-C do not state the number of generated samples, the number of independent repeats, the random seeds, or the sampling procedure used to compute the metrics. No error bars or confidence intervals are reported, yet the qualitative conclusions depend on small differences such as CharRNN FCD 0.073 versus VAE 0.099 and VAE SNN 0.626 versus CharRNN 0.602; these differences could easily arise from sampling noise. Without a specified protocol and uncertainty estimates, the numbers in Table I cannot substantiate the paper's central claims.
  2. [§IV-A, Table I; §III-B] For the overlapping metrics, the reported values match the published MOSES benchmark table digit-for-digit (for example, CharRNN validity 0.975, FCD 0.073; VAE SNN 0.626; JTN-VAE validity 1.0, FCD 0.395). The paper cites MOSES as reference [16] but never states that its results are taken from that benchmark; Section IV-A instead asserts an original comprehensive evaluation. If the numbers are reproduced from [16], the paper must say so and provide the corresponding citation for each table entry. If they are new results, then the training and evaluation details in Section III-B and III-C are insufficient to reproduce them. Either way, the empirical basis for the central claim is not established in this manuscript.
  3. [§IV-C] The property prediction results (MSE 4.37°C, R² 0.91) are presented without any description of the dataset, the train/test split, the molecular representation used for the random forest, or how the boiling-point labels were obtained. This section is also disconnected from the generative model evaluation; it does not show how property prediction complements generation, despite the introduction promising that connection. As written, the reported performance cannot be assessed or reproduced.
  4. [§IV-B, Fig. 1] The text in Section IV-B refers to Figure 1 as illustrating property distributions, and the caption appears after the discussion, but no actual figure is present in the manuscript—only two sub-caption fragments. Consequently, the Wasserstein-distance results for logP, SA, QED, and molecular weight that support the property-preservation claims for VAE and CharRNN are unverifiable.
  5. [§V-A] The 'no free lunch' argument is invoked with a placeholder citation '[ ?]' and is asserted rather than derived. The observed trade-offs are qualitative readings of a single benchmark run, not demonstrated violations of any formal theorem, and the claim that the results 'align with' a no-free-lunch theorem is not substantiated. Either provide a concrete argument connecting the observed metric trade-offs to a formal no-free-lunch statement, or remove the reference.
minor comments (5)
  1. [Abstract] There is a typo in 'be nchmarking' and inconsistent hyphenation of 'trade-offs' (also written 'tradeoffs' elsewhere); the paper should be carefully proofread.
  2. [§IV-B] The paragraph beginning 'SNN Molecular WeightScaff Novelty' contains garbled column headers and fragmented text, making it unreadable; the intended table columns should be restored.
  3. [§III-B] Training details are incomplete: no number of epochs, batch size, optimizer, learning rate, or hardware are given for the generative models, and the KL annealing schedule is described only as 'linear' with no endpoints.
  4. [References] Reference [48] is a paper on machine-learned force fields and is cited in support of the claim that hierarchical generation parallels medicinal chemistry thinking; this citation does not appear relevant to the claim and should be replaced or removed.
  5. [§IV-B] The caption for Figure 1 says 'Lower values indicate better distribution matching,' but the text in Section IV-B describes models that 'captured distributions' without consistently referencing the direction; clarify the sign convention in the text as well.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the paper reports an external benchmark evaluation against MOSES and does not derive its conclusions from fitted parameters or self-citations.

full rationale

The paper's central claim—that different generative architectures exhibit complementary strengths—rests on comparing model outputs against the external MOSES benchmark and its associated metric suite. Section III-C defines the metrics (validity, FCD, SNN, scaffold similarity, property Wasserstein distances) as external evaluation tools, and Section IV-A reports the resulting table. None of these quantities is obtained by fitting a parameter to a target and then predicting that same target; the models are trained on the MOSES training split and evaluated on held-out test and scaffold-test splits, which is a standard, non-circular protocol. The manuscript cites the original MOSES paper [16] as the source of the platform and metrics, and the reported comparisons are against that external reference rather than against a quantity defined by the paper itself. There are no self-citations by the authors, no imported uniqueness theorem, and no ansatz that is smuggled in via citation. The missing placeholder reference in Section V-A, the absence of a detailed training protocol, and the lack of error bars are completeness and reproducibility concerns, but they are not circularity: they do not make the conclusion equivalent to its inputs by construction. Accordingly, no circular step is identified and the score is set to 0.

Assumptions & free parameters 3 free parameters · 3 assumptions · 0 invented entities

No free parameters are fitted to the benchmark results; the listed parameters are hand-chosen hyperparameters that affect the reported numbers. The main additional assumptions are the validity of the MOSES metrics and a hand-wavy no-free-lunch argument.

free parameters (3)
  • CharRNN sampling temperature = 1.0
    Set by hand in Section III-B.1; affects the validity/novelty balance and thus the reported trade-offs.
  • VAE latent dimension = 128
    Chosen in Section III-B.2; influences the smoothness and conservatism of the latent space.
  • VAE KL annealing schedule = linear over 10 epochs
    Mentioned in Section III-B.2; controls the balance between reconstruction and prior, affecting novelty.
assumptions (3)
  • domain assumption MOSES metrics are valid proxies for molecular generation quality
    The paper's comparative conclusions in Section IV-A rest on these metrics being meaningful for drug discovery; Section III-C defines them but does not justify their sufficiency.
  • standard math RDKit validity and filter checks are correct
    Validity and filters are computed with RDKit (Section III-C.1); assumed to be reliable.
  • ad hoc to paper No-free-lunch argument applies to molecular generation
    Section V-A invokes the theorem to justify that no model excels across all metrics, but the citation is missing and the inference is not derived.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Comprehensive Benchmarking Platform for Deep Generative Models in Molecular Design." pith.science (2026). https://pith.science/paper/MTOAE66G

@misc{pith2026250512848,
  author       = {Pith},
  title        = {Pith review of: A Comprehensive Benchmarking Platform for Deep Generative Models in Molecular Design},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MTOAE66G}},
  note         = {Machine review of arXiv:2505.12848}
}
read the original abstract

The development of novel pharmaceuticals represents a significant challenge in modern science, with substantial costs and time investments. Deep generative models have emerged as promising tools for accelerating drug discovery by efficiently exploring the vast chemical space. However, this rapidly evolving field lacks standardized evaluation protocols, impeding fair comparison between approaches. This research presents an extensive analysis of the Molecular Sets (MOSES) platform, a comprehensive benchmarking framework designed to standardize evaluation of deep generative models in molecular design. Through rigorous assessment of multiple generative architectures, including recurrent neural networks, variational autoencoders, and generative adversarial networks, we examine their capabilities in generating valid, unique, and novel molecular structures while maintaining specific chemical properties. Our findings reveal that different architectures exhibit complementary strengths across various metrics, highlighting the complex trade-offs between exploration and exploitation in chemical space. This study provides detailed insights into the current state of the art in molecular generation and establishes a foundation for future advancements in AI-driven drug discovery.

Figures

Figures reproduced from arXiv: 2505.12848 by the authors.

Figure 1
Figure 1. Wasserstein-1 distances between property distributions of generated molecules and test set molecules. Lower values indicate better distribution matching [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

57 extracted references · 50 canonical work pages

  1. [16]

    Molecular Sets (MOSES): A B enchmarking Platform for Molecular Generation Models,

    D. Polykovskiy, A. Zhebrak, B. Sanchez -Lengeling, S. Golovanov, O. Tatanov, S. Belyaev, R. Kurbanov, A. Artamonov, V. Aladinskiy, M. Veselov, A. Kadurin, S. Johansson, H. Chen, S. Nikolenko, A. Aspuru-Guzik, and A. Zhavoronkov, “Molecular Sets (MOSES): A B enchmarking Platform for Molecular Generation Models,” Frontiers in Pharmacology , vol. 11, p. 565644, 2020

    work page 2020

  2. [1]

    Innovation in the pharmaceutical industry: New estimates of r&d costs,

    J. A. DiMasi, H. G. Grabowski, and R. W. Hansen, “Innovation in the pharmaceutical industry: New estimates of r&d costs,” Journal of Health Economics, vol. 47, pp. 20–33, 2016

    work page 2016

  3. [2]

    An analysis of the attrition of drug candidates from four major pharmaceutical companies,

    M. J. Waring, J. Arrowsmith, A. R. Leach, P . D. Leeson, S. Mandrell, R. M. Owen, G. Pairaudeau, W. D. Pennie, S. D. Pickett, J. Wang et al. , “An analysis of the attrition of drug candidates from four major pharmaceutical companies,” Nature Reviews Drug Discovery, vol. 14, no. 7, pp. 475–486, 2015

    work page 2015

  4. [3]

    2015 fda drug approvals,

    A. Mullard, “2015 fda drug approvals,” Nature Reviews Drug Discovery, vol. 15, no. 2, pp. 73–76, 2016

    work page 2015

  5. [4]

    Estimation of the size of drug-like chemical space based on gdb -17 data,

    P . G. Polishchuk, T. I. Madzhidov, and A. Varnek, “Estimation of the size of drug-like chemical space based on gdb -17 data,” Journal of ComputerAided Molecular Design, vol. 27, no. 8, pp. 675–679, 2013

    work page 2013

  6. [5]

    Rethinking drug design in the artificial intelligence era,

    G. Schneider, “Rethinking drug design in the artificial intelligence era,” Nature Reviews Drug Discovery, vol. 19, no. 5, pp. 353–364, 2020

    work page 2020

  7. [6]

    Applications of machine learning in drug discovery and development,

    J. Vamathevan, D. Clark, P . Czodrowski, I. Dunham, E. Ferran, G. Lee, B. Li, A. Madabhushi, P . Shah, M. Spitzer et al. , “Applications of machine learning in drug discovery and development,” Nature Reviews Drug Discovery, vol. 18, no. 6, pp. 463–477, 2019

    work page 2019

  8. [7]

    The rise of deep learning in drug discovery,

    H. Chen, O. Engkvist, Y . Wang, M. Olivecrona, and T. Blaschke, “The rise of deep learning in drug discovery,” Drug Discovery Today, vol. 23, no. 6, pp. 1241–1250, 2018

    work page 2018

Show all 57 references
  1. [8]

    Deep learning for molecular design —a review of the state of the art,

    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, vol. 4, no. 4, pp. 828–849, 2019

    2019

  2. [9]

    The´ power of deep learning to ligand -based novel drug discovery,

    J. A. d. Los Ramos, M. Schmidt, M. Stumpp, and M. Kowalewski, “The´ power of deep learning to ligand -based novel drug discovery,” Expert Opinion on Drug Discovery, vol. 14, no. 9, pp. 819–832, 2019

    2019

  3. [10]

    Inverse molecular design using machine learning: Generative models for matter engineering,

    B. Sanchez -Lengeling and A. Aspuru -Guzik, “Inverse molecular design using machine learning: Generative models for matter engineering,” Science, vol. 361, no. 6400, pp. 360–365, 2018

    2018

  4. [11]

    Applications of deep learning in molecule generation and molecular property prediction,

    W. P . Walters and R. Barzilay, “Applications of deep learning in molecule generation and molecular property prediction,” Accounts of Chemical Research, vol. 54, no. 2, pp. 263–270, 2020

    2020

  5. [12]

    A deep learning approach to antibiotic discovery,

    J. M. Stokes, K. Yang, K. Swanson, W. Jin, A. Cubillos-Ruiz, N. M. Donghia, C. R. MacNair, S. French, L. A. Carfrae, Z. Bloom-Ackermann et al., “A deep learning approach to antibiotic discovery,” Cell, vol. 180, no. 4, pp. 688– 702, 2020

    2020

  6. [13]

    Deep learning enables rapid identification of potent ddr1 kinase inhibitors,

    A. Zhavoronkov, Y . A. Ivanenkov, A. Aliper, M. S. Veselov, V. A. Aladinskiy, A. V. Aladinskaya, V. A. Terentiev, D. A. Polykovskiy, M. D. Kuznetsov, A. Asadulaev et al., “Deep learning enables rapid identification of potent ddr1 kinase inhibitors,” Nature Biotechnology, vol. ...

    2019

  7. [14]

    Guacamol: Benchmarking models for de novo molecular design,

    N. Brown, M. Fiscato, M. H. Segler, and A. C. Vaucher, “Guacamol: Benchmarking models for de novo molecular design,” Journal of Chemical Information and Modeling, vol. 59, no. 3, pp. 1096–1108, 2019

    2019

  8. [15]

    De novo molecular design and generative models,

    J. Meyers, B. Fabian, and N. Brown, “De novo molecular design and generative models,” Drug Discovery Today , vol. 26, no. 11, pp. 2707 – 2715, 2021

    2021

  9. [17]

    A practical overview of quantitative structure-activity relationship,

    C. Nantasenamat, C. Isarankura-Na-Ayudhya, and V. Prachayasittikul, “A practical overview of quantitative structure-activity relationship,” EXCLI Journal, vol. 9, pp. 1–23, 2010

    2010

  10. [18]

    Smiles, a chemical language and information system. 1. introduction to methodology and encoding rules,

    D. Weininger, “Smiles, a chemical language and information system. 1. introduction to methodology and encoding rules,” Journal of Chemical Information and Computer Sciences, vol. 28, no. 1, pp. 31–36, 1988

    1988

  11. [19]

    Convolutional networks on graphs for learning molecular fingerprints,

    D. K. Duvenaud, D. Maclaurin, J. Iparraguirre, R. Bombarell, T. Hirzel, A. Aspuru-Guzik, and R. P . Adams, “Convolutional networks on graphs for learning molecular fingerprints,” Advances in Neural Information Processing Systems, vol. 28, 2015

    2015

  12. [20]

    Extended-connectivity fingerprints,

    D. Rogers and M. Hahn, “Extended-connectivity fingerprints,” Journal of Chemical Information and Modeling, vol. 50, no. 5, pp. 742–754, 2010

    2010

  13. [21]

    3d molecular representations based on the wave transform for convolutional neural networks,

    D. Kuzminykh, D. Polykovskiy, A. Kadurin, A. Zhebrak, I. Baskov, S. Nikolenko, R. Shayakhmetov, and A. Zhavoronkov, “3d molecular representations based on the wave transform for convolutional neural networks,” Molecular Pharmaceutics , vol. 15, no. 10, pp. 4378 –4385, 2018

    2018

  14. [22]

    Generating focused molecule libraries for drug discovery with recurrent neural networks,

    M. H. Segler, T. Kogej, C. Tyrchan, and M. P . Waller, “Generating focused molecule libraries for drug discovery with recurrent neural networks,” ACS Central Science, vol. 4, no. 1, pp. 120–131, 2018

    2018

  15. [23]

    Generative recurrent networks for de novo drug design,

    A. Gupta, A. T. Muller, B. J. Huisman, J. A. Fuchs, P . Schneider, and¨ G. Schneider, “Generative recurrent networks for de novo drug design,” Molecular Informatics, vol. 37, no. 1-2, p. 1700111, 2018

    2018

  16. [24]

    Automatic chemical design using a data-driven continuous representation of molecules,

    R. Gomez-Bombarelli, J. N. Wei, D. Duvenaud, J. M. Hern´ andez-´ Lobato, B. Sanchez-Lengeling, D. Sheberla, J. Aguilera-Iparraguirre,´ T. D. Hirzel, R. P . Adams, and A. Aspuru-Guzik, “Automatic chemical design using a data-driven continuous representation of molecules,” ACS C...

    2018

  17. [25]

    Auto-encoding variational bayes,

    D. P . Kingma and M. Welling, “Auto-encoding variational bayes,” arXiv preprint arXiv:1312.6114, 2013

    2013 arXiv

  18. [26]

    De novo design of new chemical entities with reinvent,

    N. De Cao and T. Kipf, “De novo design of new chemical entities with reinvent,” ACS Central Science, vol. 4, no. 11, pp. 1708–1719, 2018

    2018

  19. [27]

    A de novo molecular generation method using latent vector based generative adversarial network,

    O. Prykhodko, S. V. Johansson, P .-C. Kotsias, J. Arus-Pous, E. J. Bjer-´ rum, O. Engkvist, and H. Chen, “A de novo molecular generation method using latent vector based generative adversarial network,” Journal of Cheminformatics, vol. 11, no. 1, pp. 1–13, 2019

    2019

  20. [28]

    Generative adversarial nets,

    I. Goodfellow, J. Pouget-Abadie, M. Mirza, B. Xu, D. Warde-Farley, S. Ozair, A. Courville, and Y . Bengio, “Generative adversarial nets,” Advances in Neural Information Processing Systems, vol. 27, pp. 2672– 2680, 2014

    2014

  21. [29]

    Graphnvp: An invertible flow model for generating molecular graphs,

    K. Madhawa, K. Ishiguro, K. Nakago, and M. Abe, “Graphnvp: An invertible flow model for generating molecular graphs,” arXiv preprint arXiv:1905.11600, 2019

    1905 arXiv

  22. [30]

    Mermaid: Metaphors for molecular generation,

    C. Zang and F. Wang, “Mermaid: Metaphors for molecular generation,” arXiv preprint arXiv:2004.08047, 2020

    2004

  23. [31]

    Molecular denovo design through deep reinforcement learning,

    M. Olivecrona, T. Blaschke, O. Engkvist, and H. Chen, “Molecular denovo design through deep reinforcement learning,” Journal of Cheminformatics, vol. 9, no. 1, pp. 1–14, 2017

    2017

  24. [32]

    Deep reinforcement learning for de novo drug design,

    M. Popova, O. Isayev, and A. Tropsha, “Deep reinforcement learning for de novo drug design,” Science Advances, vol. 4, no. 7, p. eaap7885, 2018

    2018

  25. [33]

    Smiles transformer: Pre -trained molecular fingerprint for low data drug discovery,

    S. Honda, S. Shi, and H. R. Ueda, “Smiles transformer: Pre -trained molecular fingerprint for low data drug discovery,” arXiv preprint arXiv:1911.04738, 2019

    1911 arXiv

  26. [34]

    Mol - cyclegan: a generative model for molecular optimization,

    Ł. Maziarka, A. Pocha, J. Kaczmarczyk, K. Rataj, and M. Warchoł, “Mol - cyclegan: a generative model for molecular optimization,” Journal of Cheminformatics, vol. 12, no. 1, pp. 1–18, 2020

    2020

  27. [35]

    The molecular transformer for organic reaction prediction and symbolic inference,

    R. Irwin, S. Dimitriadis, J. He, and E. J. Bjerrum, “The molecular transformer for organic reaction prediction and symbolic inference,” Journal of Chemical Information and Modeling, vol. 62, no. 13, pp. 3204– 3213, 2022

    2022

  28. [36]

    Frechet chemnet distance: A metric for generative models for molecules´ in drug discovery,

    K. Preuer, P . Renz, T. Unterthiner, S. Hochreiter, and G. Klambauer, “Frechet chemnet distance: A metric for generative models for molecules´ in drug discovery,” Journal of Chemical Information and Modeling, vol. 58, no. 9, pp. 1736–1741, 2018

    2018

  29. [37]

    Estimation of synthetic accessibility score of drug -like molecules based on molecular complexity and fragment contributions,

    P . Ertl and A. Schuffenhauer, “Estimation of synthetic accessibility score of drug -like molecules based on molecular complexity and fragment contributions,” Journal of Cheminformatics, vol. 1, no. 1, pp. 1–11, 2009

    2009

  30. [38]

    Zinc 15–ligand discovery for everyone,

    T. Sterling and J. J. Irwin, “Zinc 15–ligand discovery for everyone,” Journal of Chemical Information and Modeling , vol. 55, no. 11, pp. 2324 –2337, 2015

    2015

  31. [39]

    The pains problem in virtual screening,

    J. B. Baell and G. A. Holloway, “The pains problem in virtual screening,” Journal of Medicinal Chemistry, vol. 53, no. 7, pp. 2719–2740, 2010

    2010

  32. [40]

    The properties of known drugs. 1. molecular frameworks,

    G. W. Bemis and M. A. Murcko, “The properties of known drugs. 1. molecular frameworks,” Journal of Medicinal Chemistry, vol. 39, no. 15, pp. 2887–2893, 1996

    1996

  33. [41]

    Junction tree variational autoencoder for molecular graph generation,

    W. Jin, R. Barzilay, and T. Jaakkola, “Junction tree variational autoencoder for molecular graph generation,” International Conference on Machine Learning, pp. 2323–2332, 2018

    2018

  34. [42]

    Prediction of physicochemical parameters by atomic contributions,

    S. A. Wildman and G. M. Crippen, “Prediction of physicochemical parameters by atomic contributions,” Journal of Chemical Information and Computer Sciences, vol. 39, no. 5, pp. 868–873, 1999

    1999

  35. [43]

    Quantifying the chemical beauty of drugs,

    G. R. Bickerton, G. V. Paolini, J. Besnard, S. Muresan, and A. L. Hopkins, “Quantifying the chemical beauty of drugs,” Nature Chemistry, vol. 4, no. 2, pp. 90–98, 2012

    2012

  36. [44]

    Randomized smiles strings improve the quality of molecular generative models,

    J. Arus -Pous, T. Blaschke, S. Ulander, J. -L. Reymond, H. Chen, and´ O. Engkvist, “Randomized smiles strings improve the quality of molecular generative models,” Journal of Cheminformatics, vol. 11, no. 1, pp. 1–13, 2019

    2019

  37. [45]

    Learning con-´ tinuous and data-driven molecular descriptors by translating equivalent chemical representations,

    R. Winter, F. Montanari, F. Noe, and D.-A. Clevert, “Learning con-´ tinuous and data-driven molecular descriptors by translating equivalent chemical representations,” Chemical Science, vol. 10, no. 6, pp. 1692– 1701, 2019

    2019

  38. [46]

    Application of generative autoencoder in de novo molecular design,

    T. Blaschke, M. Olivecrona, O. Engkvist, J. Bajorath, and H. Chen, “Application of generative autoencoder in de novo molecular design,” Molecular Informatics, vol. 37, no. 1-2, p. 1700123, 2018

    2018

  39. [47]

    Generative models for molecular discovery: Recent advances and challenges,

    M. Moret, L. Friedrich, F. Grisoni, D. Merk, and G. Schneider, “Generative models for molecular discovery: Recent advances and challenges,” Wiley Interdisciplinary Reviews: Computational Molecular Science, vol. 10, no. 3, p. e1460, 2020

    2020

  40. [48]

    Machine learning force fields and coarse-grained variables in molecular dynamics: application to materials and biological systems,

    P . Maragakis, B. Alldritt, M. Ashton, S. Hennel, and J. B. Sorge, “Machine learning force fields and coarse-grained variables in molecular dynamics: application to materials and biological systems,” Journal of Chemical Theory and Computation, vol. 16, no. 8, pp. 4651–4665, 2020

    2020

  41. [49]

    Population-based de novo molecule generation, using grammatical evolution,

    N. Yoshikawa, K. Terayama, M. Sumita, T. Homma, K. Oono, and K. Tsuda, “Population-based de novo molecule generation, using grammatical evolution,” Chemistry Letters, vol. 47, no. 11, pp. 1431–1434, 2019

    2019

  42. [50]

    Autonomous discovery in the chemical sciences part i: Progress,

    C. W. Coley, N. S. Eyke, and K. F. Jensen, “Autonomous discovery in the chemical sciences part i: Progress,” Angewandte Chemie International Edition, vol. 59, no. 51, pp. 22858–22893, 2020

    2020

  43. [51]

    Deep learning for drug design: an artificial intelligence paradigm for drug discovery in the big data era,

    Y . Xu, K. Lin, S. Wang, L. Wang, C. Cai, C. Song, L. Lai, and J. Pei, “Deep learning for drug design: an artificial intelligence paradigm for drug discovery in the big data era,” Medicinal Research Reviews, vol. 41, no. 4, pp. 1734–1766, 2021

    2021

  44. [52]

    Recent applications of machine learning in drug discovery,

    J. Panteleev, H. Gao, and L. Jia, “Recent applications of machine learning in drug discovery,” Bioorganic & Medicinal Chemistry Letters, vol. 28, no. 17, pp. 2807–2815, 2018

    2018

  45. [53]

    A practical guide to molecular generative models,

    S. R. Atance, J. V. Diez, and O. Engkvist, “A practical guide to molecular generative models,” Journal of Cheminformatics, vol. 14, no. 1, pp. 1– 19, 2022

    2022

  46. [54]

    Exploration strategies for discovery of interacting multi-target drugs,

    J. Jimenez-Luna, M. Skalic, N. Weskamp, and G. Schneider, “Exploration strategies for discovery of interacting multi-target drugs,” Drug Discovery Today, vol. 25, no. 2, pp. 261–267, 2020

    2020

  47. [55]

    Artificial intelligence in drug discovery,

    M. A. Sellwood, M. Ahmed, M. H. Segler, and N. Brown, “Artificial intelligence in drug discovery,” Future Medicinal Chemistry, vol. 10, no. 17, pp. 2025–2028, 2018

    2025

  48. [56]

    Molgpt: Molecular generation using a transformer -decoder model,

    M. Thomas, R. T. Smith, N. M. O’Boyle, C. de Graaf, and A. Bender, “Molgpt: Molecular generation using a transformer -decoder model,” Journal of Cheminformatics, vol. 14, no. 1, pp. 1–19, 2022

    2022

  49. [57]

    Diffdock:¨ Diffusion steps, twists, and turns for molecular docking,

    G. Corso, H. Stark, B. Jing, R. Barzilay, and T. Jaakkola, “Diffdock:¨ Diffusion steps, twists, and turns for molecular docking,” arXiv preprint arXiv:2210.01776, 2022

    2022 arXiv

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.