Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

Enhancing Drug Discovery: Quantum Machine Learning for QSAR Prediction with Incomplete Data

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read The paper claims that a hybrid quantum classifier generalizes better than a simple classical neural network on QSAR drug-discovery data when features and training samples are scarce.

desk verdict The paper's central quantum-advantage claim is undercut by a parameter-count error: by its own formulas the PQC has 50–65% more parameters than the tiny classical MLP, not 82% fewer, so the comparison does not control for capacity. read the letter →

arxiv 2501.13395 v1 pith:ZLG6C7AI submitted 2025-01-23 quant-ph

Wei-Yin Chiang , Po-Yu Kao , Tzu-Lan Yeh , Ya-Chu Yang , Yen-Chu Lin , Alex Zhavoronkov This is my paper · ORCID
classification quant-ph
keywords QSARquantummachinelearningparameterizedcircuitPCAfeatureselectiongeneralizationsmalldatadrugdiscoverymolecularembedding
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

This paper tries to establish that a quantum classifier generalizes better than a simple classical neural network on drug-discovery QSAR tasks when the data are incomplete: few features and few training samples. Using Morgan fingerprints or ImageMol embeddings, then PCA to cut the feature count, the authors compare a parameterized quantum circuit with amplitude embedding to a three-layer perceptron on the BACE, BBBP, and HIV datasets. They report that the quantum classifier outperforms the classical one at n=3 (six features) on BACE by about eight percentage points, and that the advantage persists across embedding methods and appears in the small-training-sample regime at n=4. The authors attribute this to the quantum model's robustness to feature and sample reduction, and to a tighter generalization bound tied to a smaller VC-dimension growth for circuits. If correct, this suggests PQC-based classifiers are a practical option for medical datasets where missing features and scarce samples are the norm.

What carries the argument

The central object is the parameterized quantum circuit (PQC) classifier: n qubits, amplitude embedding of 2n PCA-selected features, two strongly entangled layers each with 3n rotation angles, and measurement of all qubits followed by a linear regression with a zero threshold, giving 7n trainable parameters. The classical comparator is a three-layer MLP of shape N×2×1 with 2(N+1) parameters. The argument runs on the contrast in how the two models' generalization error scales with parameters: the paper invokes classical VC dimension proportional to ρ for MLPs and quantum VC dimension proportional to Γ log₂Γ for circuits, so that the quantum model's complexity grows more slowly and its generalization bound is tighter, which the authors use to explain the observed robustness to reduced features and samples.

What would settle it

Train the classical MLP on BACE Morgan features with a hidden layer of 8 or 16 neurons (or with dropout/weight decay) at n=3 and compare accuracy; if the classical model reaches or exceeds 0.75, the quantum advantage disappears. Alternatively, run the paper's own 5×20 protocol with a paired significance test: if the 0.75 vs 0.69 difference has p>0.05, the advantage is not established.

Watch

Extended reading notes

Core claim

The central claim is that a hybrid quantum-classical classifier, built from a parameterized quantum circuit with amplitude encoding and two strongly entangling layers, outperforms a classical three-layer perceptron with comparable structure when the input dimension and training set are small. The evidence is accuracy comparisons on QSAR datasets after PCA feature selection: on BACE with Morgan fingerprints, the quantum classifier reaches 0.75 accuracy at n=3 versus 0.69 for the classical model, and at n=4 the two tie while the quantum model is said to use fewer trainable parameters; on BBBP and HIV the advantage at n=3 is about 1–2 percentage points. The authors interpret the pattern as a difference in generalization power: the classical model's accuracy drops sharply as features and samples shrink, while the quantum model degrades more slowly, and at equal training loss the quantum model has higher test accuracy. They attribute this to the VC dimension of quantum circuits scaling more slowly with parameters than that of multilayer perceptrons.

Load-bearing premise

The load-bearing premise is that the three-layer MLP with a two-neuron hidden layer is the right classical baseline; if a stronger classical model—wider, regularized, or with better feature handling—matches the quantum accuracy on six features, the claimed quantum advantage in generalization does not hold.

Editorial extensions

If this is right

  • Drug-discovery pipelines with sparse or incomplete descriptor tables can use PQC-based classifiers to maintain accuracy where classical MLPs degrade.
  • The advantage is not tied to one embedding: it appears for both Morgan fingerprints and ImageMol image embeddings, so it is a property of the classifier rather than the molecular representation.
  • Quantum classifiers are more robust to feature-dimensionality reduction, so they are better suited to PCA-compressed or missing-feature regimes.
  • On larger feature sets (n=8), the classical MLP overtakes the quantum classifier, so the benefit is confined to the low-dimensional regime.
  • The VC-dimension argument suggests PQC models as a general recipe for small-sample classification tasks beyond QSAR, wherever classical models overfit on few examples.

Reading between the lines

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

  • The comparison could be sharpened by testing the classical MLP with wider hidden layers or explicit regularization; if the gap closes, the quantum advantage would look like a capacity or regularization effect rather than a quantum one.
  • A significance test (e.g., paired bootstrap over the 5×20 training runs) would tell whether the reported 1–8% differences are reproducible; the paper does not report one.
  • The paper's parameter-count comparison (quantum 7n vs classical 4n+2) contradicts the stated 82% at n=4; resolving this arithmetic is a prerequisite for the expressive-power argument.
  • The robustness claim could be tested on other open QSAR sets with controlled missing-feature masks, beyond the three datasets used here.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes a workflow for QSAR classification using classical and parameterized-quantum-circuit (PQC) classifiers on BACE, BBBP, and HIV datasets. Molecular features are embedded with Morgan fingerprints or ImageMol, reduced by PCA to 2n features, and then classified by a three-layer MLP or a two-layer strongly entangling PQC with n qubits. The central claim is that the quantum classifier outperforms the classical one when the number of selected features and training samples is small, with Table 1 reporting accuracies of 0.75 versus 0.69 (BACE, Morgan, n=3) and 0.70 versus 0.65 (BACE, ImageMol, n=3). The paper also examines training-loss/testing-accuracy relationships, cluster-based small-sample behavior, and the generality of the advantage on BBBP and HIV.

Significance. If the claimed small-feature, small-sample quantum advantage were robust, it would be a practically relevant result for QSAR with incomplete data and would add to the growing evidence on PQC generalization. The paper's strengths include a clear and simple architecture comparison, a repeated train-test evaluation protocol, and use of two distinct molecular embeddings. However, the significance is materially weakened by an internal contradiction in the parameter-count comparison (the quantum classifier actually has more trainable parameters under the paper's own formulas), by the absence of significance tests for the small accuracy differences, and by the authors' own admission that the external-dataset advantages are not significant. These issues must be resolved before the central claim can be considered established.

major comments (4)
  1. [Methods (Classifier); Results (Fig. 3(b)); Discussion] The paper states that the classical classifier has 2(N+1) trainable parameters with N=2n and that the quantum classifier has 7n parameters (two strongly entangling layers with 3n variables each, plus a linear readout), and then claims the quantum classifier needs only 82% of the classical parameters. Substituting N=2n gives classical parameters 4n+2 and quantum parameters 7n, so the quantum classifier has 7n/(4n+2) ≈ 1.50, 1.56, and 1.65 times more parameters for n=3, 4, and 8, respectively. The '82%' and '18% fewer' statements are therefore contradicted by the paper's own architecture; the comparison is not capacity-controlled as claimed.
  2. [Results (Tables 1, and Figs. 4 and 6)] No significance tests or effect-size measures are reported anywhere, and the only numerical uncertainties are the ± values in Table 1. For example, the BACE/MGFP n=3 difference (0.75 versus 0.69) and the BACE/IMGMOL n=3 difference (0.70 versus 0.65) are within two standard deviations; without paired tests or confidence intervals, these gaps cannot be distinguished from run-to-run noise. Figures 4 and 6 additionally show no error bars, even though they directly support the central small-training-sample claim.
  3. [Results (Generality of quantum advantage on other datasets); Discussion; Conclusion] The external-dataset results do not support the conclusion that quantum advantages generalize. The paper reports advantages of roughly 1% (BBBP) and 2% (HIV) with Morgan embedding, states in the Discussion that these are 'not significant from the statistical results,' and notes that ImageMol embedding removes the advantage. Yet the Conclusion says the quantum classifier outperforms the classical one on 'various datasets.' The conclusion overstates the evidence presented in the manuscript.
  4. [Discussion (VC dimension paragraph)] The VC-dimension explanation is not derived for the specific circuits used here; it relies on scaling relations cited from Refs. [30] and [31]. Since the quantum classifier in this paper actually has more trainable parameters than the classical MLP, the claim that the quantum model has a tighter generalization bound is not established by the cited material, and the argument does not cure the capacity mismatch identified above.
minor comments (5)
  1. [Results (Table 1 discussion)] The sentence 'the classical classifier performs better when data embedded by MGFP with 28 (n = 8) features' appears to contain a typo: 2n=16 features at n=8, not 28.
  2. [Figure 1 caption] The caption contains 'PrincipleComponent Analysis' and should read 'Principal Component Analysis.'
  3. [Figure 8 caption] The caption says 'IMAGMOL embedding' but the method is 'ImageMol'; please correct the spelling.
  4. [Appendix A] The sentence 'the classification accuracy of the quantum classifier is wore in most cases' contains a typo; 'wore' should be 'worse.'
  5. [General] The paper does not state whether code or data are available; providing the training/evaluation scripts would improve reproducibility of the reported accuracies and error estimates.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the quantum-advantage claim is an empirical benchmark result, not a derived quantity, and the only self-citation is a background citation that is not load-bearing.

full rationale

The paper's central claim is an empirical comparison: classifiers are trained on fixed embeddings and PCA-selected features, and test-set accuracies are reported in Table 1 and Figures 3-8. No parameter is fitted to the reported accuracy and then renamed a prediction; no result is defined in terms of the claim it is supposed to support. The only potentially load-bearing theoretical explanation is the VC-dimension comparison in the Discussion, but it is explicitly cited to external sources (Bartlett & Maass, Ref. 30, and Caro, Ref. 31) and is used as a post hoc explanation rather than as the source of the empirical result. The one self-citation (Kao et al., Ref. 21, cited in the Introduction for quantum generative adversarial networks in drug discovery) is background motivation and is not load-bearing. The parameter-count contradiction noted by the reader (7n quantum versus 4n+2 classical parameters, contradicting the '82%' claim) is an internal-consistency/correctness issue, not a circularity issue, because the experimental comparison itself does not derive the advantage from that count. The paper's own Discussion even concedes that the advantage is not significant on BBBP/HIV and is partition-dependent, which is inconsistent with a claim that is forced by construction. Accordingly, no circular step can be exhibited under the required quote-and-reduction standard.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The paper is an empirical benchmark with no derived formula, so its auxiliary assumptions are standard ML and QML background. The main concerns are the hand-chosen classical baseline, the post hoc definition of the low-dimensional regime, and the absence of code, none of which introduce invented entities.

free parameters (4)
  • n (qubit count and log2 feature count) = 2, 3, 4, 8
    The central claim is conditioned on small n; the paper scans these values and declares the n=3 to n=4 range the advantageous regime.
  • Number of strongly entangling layers = 2
    Chosen by hand for the ansatz; no ablation is shown.
  • Classical MLP hidden width = 2
    The N x 2 x 1 perceptron is a deliberately minimal baseline; its weakness directly affects the comparison.
  • Training epochs and repetitions = 100 epochs, 20 trainings, 5 resamples
    Protocol details chosen by the authors; the paper takes the best accuracy among 100 epochs, which favors memorization.
assumptions (3)
  • standard math Amplitude embedding can faithfully encode 2n real features into n qubits
    Assumes feature vectors can be normalized and loaded into quantum states; standard but not validated for these chemical embeddings.
  • domain assumption VC dimension of classical MLP scales as rho^2 and VC dimension of PQC scales as Gamma log^2 Gamma
    The generalization explanation relies on Refs 30 and 31; no derivation is given in this paper.
  • domain assumption Undersampling BBBP and HIV data preserves the comparison
    Undersampling is applied to address class imbalance but is not analyzed for its effect on the classifier comparison.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Enhancing Drug Discovery: Quantum Machine Learning for QSAR Prediction with Incomplete Data." pith.science (2026). https://pith.science/paper/ZLG6C7AI

@misc{pith2026250113395,
  author       = {Pith},
  title        = {Pith review of: Enhancing Drug Discovery: Quantum Machine Learning for QSAR Prediction with Incomplete Data},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ZLG6C7AI}},
  note         = {Machine review of arXiv:2501.13395}
}
read the original abstract

Qualitative structure-activity relationship (QSAR) is important for drug discovery and offers valuable insights into the biological interactions of potential drug candidates. It has been demonstrated that QSAR can be accurately predicted by machine learning. However, data with poor quality and limited availability are always the most common and critical issues for medical-related applications for machine learning. In this manuscript, we aim to discuss the performance of classical and quantum classifiers in QSAR prediction and attempt to demonstrate the quantum advantages in the generalization power of the quantum classifier under conditions of limited data availability and a reduced number of features. By applying different data embedding methods followed by feature selection through principal component analysis (PCA), we find that the quantum classifier outperforms the classical one when a small number of features are selected and the number of training samples is limited. The generality of quantum advantages in other open datasets is also explored.

Figures

Figures reproduced from arXiv: 2501.13395 by the authors.

Figure 1
Figure 1. The proposed workflow. The inhibitor structure(SMILES) is embedded by Morgan fingerprint or ImageMol followed by feature selection (PrincipleComponent Analysis, PCA). 2n selected features are fed into classical and quantum classifiers. Where n is the number of qubits in the quantum classifier. Molecular Embedding In this section, we will provide a concise overview of the two distinct types of molecular embeddings ut… view at source ↗
Figure 2
Figure 2. The architectural of classifiers. (a) In classical classifier, the classical multi-layer perceptron (MLP) has an input layer with 2 n neurons, a hidden layer with two nodes, and an output node. (b) The quantum classifier consists of an amplitude encoding layer to encode input with the dimension of 2n , two strongly entangling layers followed by the measurements to all qubits. The measured results are then aggregated… view at source ↗
Figure 3
Figure 3. (a) The trend in accuracy with varying numbers of features. The change in accuracy is highly dependent on the number of features in the classical classifier. (b) and (c) show the relationships between training loss and testing accuracy for n = 4 and n = 3 respectively. From these training loss and testing accuracy results, it is evident that the quantum classifier exhibits superior generalization power compared to t… view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: (a) and (b) are the dependencies of accuracies on the number of training data when n = 8 and n = 4, respectively. The figure shows that when n = 4, the quantum advantage is found at the small number of training data regimes, while the classical classifier still outperf…
Figure 5
Figure 5. Figure 5: To test the generalization ability of classical and quantum models, cluster29 of embedded molecules was conducted, and 1 to 7 data points from each larger cluster were selected as training data. (embedding method: MGFP) [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: The comparisons of the testing accuracies when different numbers of samples picked from each cluster are used for training. In the scenario of low-dimensional features, n = 3, the testing accuracies are significantly higher in the quantum classifier [PITH_FULL_IMAGE:f…
Figure 7
Figure 7. Figure 7: Performances of the classical and quantum classifiers: The histogram depicts the prediction accuracies under MGFP embedding for both n = 3 and n = 8. From the figure, it is evident that the quantum classifier performs slightly better than the classical one when n is sm…
Figure 8
Figure 8. Figure 8: Performances of the classical/quantum classifiers. The histogram of the prediction accuracies under IMAGMOL embedding when n = 3 and n = 8. Data embedded by IMGMOL enhances the performances of classical classifiers but quantum classifiers. This improvement vanishes the…
Figure 9
Figure 9. Figure 9: QSAR prediction on MGFP embedded BBBP(top) and HIV(bottom) datasets from classical(blue) and quantum(green) classifiers. The results in different columns are obtained from different data partitions. Quantum classifiers outperform the classical ones in low-dimensional f…
Figure 10
Figure 10. Figure 10: QSAR prediction on IMGMOL embedded BBBP(top) and HIV(bottom) datasets from classical(blue) and quantum(green) classifiers. The results in different columns are obtained from different data partitions. Quantum classifiers outperform the classical ones in low-dimensiona…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. $\mathtt{Q^2SAR}$: overcoming classical bottlenecks in drug discovery via quantum multiple kernel learning

    quant-ph 2026-07 conditional novelty 4.0 of 10

    Simulated QMKL-SVM reaches AUC ~0.87–0.90 on DYRK1A QSAR after PCA reduction, beating a same-space gradient-boosting baseline of ~0.80.

Reference graph

Works this paper leans on

36 extracted references · 30 canonical work pages · cited by 1 Pith paper

  1. [30]

    Bartlett, P. L. & Maass, W. Vapnik-chervonenkis dimension of neural nets. The handbook brain theory neural networks 1188–1192 (2003)

    work page 2003

  2. [31]

    Caro, M. C. Quantum Learning Theory. Ph.D. thesis, Technische Universität München (2022)

    work page 2022

  3. [1]

    & Fujita, T

    Hansch, C. & Fujita, T. p-σ-π analysis. a method for the correlation of biological activity and chemical structure. J. Am. Chem. Soc. 86, 1616–1626 (1964)

    work page 1964

  4. [2]

    Boosting support vector regression in qsar studies of bioactivities of chemical compounds

    Zhou, Y .-P.et al. Boosting support vector regression in qsar studies of bioactivities of chemical compounds. Eur. journal pharmaceutical sciences 28, 344–353 (2006)

    work page 2006

  5. [3]

    Pourbasheer, E., Riahi, S., Ganjali, M. R. & Norouzi, P. Application of genetic algorithm-support vector machine (ga-svm) for prediction of bk-channels activity. Eur. journal medicinal chemistry 44, 5023–5028 (2009)

    work page 2009

  6. [4]

    Polishchuk, P. G.et al. Application of random forest approach to qsar prediction of aquatic toxicity.J. chemical information modeling 49, 2481–2488 (2009)

    work page 2009

  7. [5]

    E., Jaitly, N

    Dahl, G. E., Jaitly, N. & Salakhutdinov, R. Multi-task neural networks for qsar predictions.arXiv preprint arXiv:1406.1231 (2014)

    arXiv 2014

  8. [6]

    & Yoon, S

    Kwon, S., Bae, H., Jo, J. & Yoon, S. Comprehensive ensemble in qsar prediction for drug discovery. BMC bioinformatics 20, 1–12 (2019)

    work page 2019

Show all 36 references
  1. [7]

    & Ganjali, M

    Pourbasheer, E., Aalizadeh, R. & Ganjali, M. R. Qsar study of ck2 inhibitors by ga-mlr and ga-svm methods. Arab. J. Chem. 12, 2141–2149 (2019)

    2019

  2. [8]

    & Wang, B

    Hu, S., Chen, P., Gu, P. & Wang, B. A deep learning-based chemical system for qsar prediction. IEEE journal biomedical health informatics 24, 3020–3028 (2020)

    2020

  3. [9]

    C., Joe, D., Woo, Y ., Kim, Y

    Kim, B. C., Joe, D., Woo, Y ., Kim, Y . & Yoon, G. Extension of pqsar: Ensemble model generated by random forest and partial least squares regressions. IEEE Access 8, 180087–180099 (2020)

    2020

  4. [10]

    & Fiorentini, M

    Benedetti, M., Lloyd, E., Sack, S. & Fiorentini, M. Parameterized quantum circuits as machine learning models. Quantum Sci. Technol. 4, 043001 (2019)

    2019

  5. [11]

    Biamonte, J. et al. Quantum machine learning. Nature 549, 195–202 (2017)

    2017

  6. [12]

    L., West, M

    Tsang, S. L., West, M. T., Erfani, S. M. & Usman, M. Hybrid quantum-classical generative adversarial network for high resolution image generation. arXiv preprint arXiv:2212.11614 (2022)

    2022 arXiv

  7. [13]

    Huang, H.-L. et al. Experimental quantum generative adversarial networks for image generation. Phys. Rev. Appl. 16, 024051 (2021)

    2021

  8. [14]

    & Katouda, M

    Suzuki, T. & Katouda, M. Predicting toxicity by quantum machine learning. J. Phys. Commun. 4, 125012 (2020)

    2020

  9. [15]

    Wu, S. L. et al. Application of quantum machine learning using the quantum kernel algorithm on high energy physics analysis at the lhc. Phys. Rev. Res. 3, 033221 (2021)

    2021

  10. [16]

    Supervised learning with quantum-enhanced feature spaces

    Havlíˇcek, V .et al. Supervised learning with quantum-enhanced feature spaces. Nature 567, 209–212 (2019)

    2019

  11. [17]

    Batra, K. et al. Quantum machine learning algorithms for drug discovery applications. J. chemical information modeling 61, 2641–2647 (2021)

    2021

  12. [18]

    & Temme, K

    Liu, Y ., Arunachalam, S. & Temme, K. A rigorous and robust quantum speed-up in supervised machine learning. Nat. Phys. 17, 1013–1017 (2021)

    2021

  13. [19]

    & Tao, D

    Du, Y ., Hsieh, M.-H., Liu, T. & Tao, D. Expressive power of parametrized quantum circuits.Phys. Rev. Res. 2, 033125 (2020)

    2020

  14. [20]

    & Aspuru-Guzik, A

    Cao, Y ., Romero, J. & Aspuru-Guzik, A. Potential of quantum computing for drug discovery.IBM J. Res. Dev. 62, 6–1 (2018)

    2018

  15. [21]

    Exploring the advantages of quantum generative adversarial networks in generative chemistry

    Kao, P.-Y .et al. Exploring the advantages of quantum generative adversarial networks in generative chemistry. J. Chem. Inf. Model. (2023)

    2023

  16. [22]

    & Borondo, F

    Domingo, L., Djukic, M., Johnson, C. & Borondo, F. Hybrid quantum-classical convolutional neural networks to improve molecular protein binding affinity predictions. arXiv preprint arXiv:2301.06331 (2023)

    2023 arXiv

  17. [23]

    Sagingalieva, A. et al. Hybrid quantum neural network for drug response prediction. Cancers 15, 2705 (2023)

    2023

  18. [24]

    Hekler, E. B. et al. Why we need a small data paradigm. BMC medicine 17, 1–9 (2019)

    2019

  19. [25]

    Jolliffe, I. T. Principal component analysis for special types of data (Springer, 2002)

    2002

  20. [26]

    & Hahn, M

    Rogers, D. & Hahn, M. Extended-connectivity fingerprints. J. chemical information modeling 50, 742–754 (2010). 10/12

    2010

  21. [27]

    Zeng, X. et al. Accurate prediction of molecular properties and drug targets using a self-supervised image representation learning framework. Nat. Mach. Intell. 4, 1004–1016 (2022)

    2022

  22. [28]

    Landrum, G. et al. rdkit/rdkit: 2022_09_1b1 (q3 2022) release, DOI: 10.5281/zenodo.7179566 (2022)

    2022 doi

  23. [29]

    Unsupervised data base clustering based on daylight’s fingerprint and tanimoto similarity: A fast and automated way to cluster small and large data sets

    Butina, D. Unsupervised data base clustering based on daylight’s fingerprint and tanimoto similarity: A fast and automated way to cluster small and large data sets. J. Chem. Inf. Comput. Sci. 39, 747–750 (1999)

    1999

  24. [32]

    Mensa, S., Sahin, E., Tacchino, F., Barkoutsos, P. K. & Tavernelli, I. Quantum machine learning framework for virtual screening in drug discovery: a prospective quantum advantage. Mach. Learn. Sci. Technol. 4, 015023 (2023)

    2023

  25. [33]

    S., Saggi, M

    Bhatia, A. S., Saggi, M. K. & Kais, S. Quantum machine learning predicting adme-tox properties in drug discovery. J. Chem. Inf. Model. (2023)

    2023

  26. [34]

    & Denny, R

    Subramanian, G., Ramsundar, B., Pande, V . & Denny, R. A. Computational modeling ofβ-secretase 1 (bace-1) inhibitors using ligand based approaches. J. chemical information modeling 56, 1936–1949 (2016)

    2016

  27. [35]

    F., Teixeira, A

    Martins, I. F., Teixeira, A. L., Pinheiro, L. & Falcao, A. O. A bayesian approach to in silico blood-brain barrier penetration modeling. J. chemical information modeling 52, 1686–1697 (2012)

    2012

  28. [36]

    https://wiki.nci.nih.gov/display/NCIDTPdata/AIDS+Antiviral+Screen+Data

    Aids antiviral screen data. https://wiki.nci.nih.gov/display/NCIDTPdata/AIDS+Antiviral+Screen+Data. Accessed: 27.09.2017. Appendix A The results shown in the main text demonstrate the insignificant quantum advantage on the BBBP and HIV datasets. This insignificance results fro...

    2017

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.