Pith. sign in

REVIEW 3 major objections 6 minor 55 references

MIN: Multi-channel Interaction Network for Drug-Target Interaction with Protein Distillation

T0 review · 3 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read This paper claims that screening proteins by predicted conservation before matching them to drugs yields drug–target interaction predictions that surpass all compared baselines, with selected residues overlapping known binding pockets.

desk verdict Genuinely new DTI architecture with a real bug: the C-Score distillation's claimed DTI training signal is blocked by a hard threshold, so the mechanism doesn't do what the paper says. read the letter →

arxiv 2412.07778 v1 pith:NQO4NTKW submitted 2024-11-23 q-bio.QM cs.LG

Shuqi Li , Shufang Xie , Hongda Sun , Yuhan Chen , Tao Qin , Tianjun Ke , Rui Yan This is my paper · ORCID
classification q-bio.QMcs.LG
keywords drug-targetinteractionpredictionproteindistillationC-ScorePredictorconservationscoremultiplesequencealignmentmulti-channelcontrastivelearningbindingpocket
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 claims that drug–target interaction (DTI) prediction gets better and more interpretable when the target protein is first stripped down to its most evolutionarily conserved residues, and when the remaining representation is matched to the drug through three complementary channels that use sequence-only, structure-only, and mixed information. The proposed model, MIN, contains a C-Score Predictor that learns per-residue conservation from multiple-sequence alignments and masks non-conserved residues before encoding; a contrastive loss then aligns the sequence and structure views of each drug and target. On the DUD-E benchmark the authors report an AUC of 0.983 with a 0.5% ROC enrichment of 197.741, and on the Human benchmark an AUC of 0.984 with precision 0.983 and F1 0.965, both described as surpassing all compared baselines. In a case study, 16 of the 37 residues MIN selects overlap the true binding pocket of a PDB structure. If these claims hold, conservation-based distillation is a practical prior for locating binding sites and for improving virtual screening at low false-positive rates.

What carries the argument

The load-bearing object is the C-Score Predictor: a Transformer encoder $\varphi_c$ plus a fully connected layer $f_c$ that maps MSA row-average embeddings, residue embeddings, and positional embeddings to a per-residue conservation score $\hat{y}_s$, trained by mean-squared error against computed conservation scores. Its output is applied through Eq. (2), a hard threshold that replaces every residue whose predicted score exceeds $r_0=0.01$ with a mask embedding, so only conserved residues survive. The second mechanism is the three-channel interaction module, where three MLPs score sequence-level, structure-level, and mixed representations and a feed-forward layer fuses their outputs; contrastive InfoNCE losses align the sequence and structure representations of each drug and target. Together these components are intended to let the model focus on the conserved, likely pocket-forming residues while still exploiting both sequence and 3D contact information.

What would settle it

Check in the released training code whether the interaction-prediction loss can change the parameters of the residue-selection module through the masking step in Eq. (2); if it cannot, the claim that the module is trained with the interaction objective is not realised. Also run a direct comparison between the learned selector and a plain conservation-score cutoff on the same downstream model and the same threshold $r_0$; if AUC and enrichment are statistically indistinguishable, the learned selector adds nothing.

Watch

Extended reading notes

Core claim

MIN's central discovery claim is that treating protein conservation as a learnable screening signal, rather than as a fixed filter, improves DTI classification. The C-Score Predictor takes a protein sequence, its MSA, and positional embeddings, predicts a conservation score for every residue, and keeps only residues whose predicted score falls below the threshold $r_0=0.01$, because a lower score means higher conservation; the distilled sequence is then encoded by a Transformer while the MSA-guided contact graph is encoded by a GCN. Three interaction channels — structure-agnostic (sequence–sequence), structure-aware (graph–graph), and extended-mixture (concatenated sequence and structure) — are fused by a feed-forward layer, and InfoNCE contrastive losses pull the sequence and structure representations of the same entity together. The paper reports that MIN achieves the best AUC and enrichment numbers on DUD-E and the best AUC, precision, and F1 on Human among all baselines tested, and that ablating the C-Score Predictor produces the largest single-component drop in AUC (from 0.983 to 0.968). The case study on PDB entry 2vjx is offered as evidence that the residues MIN keeps are biologically meaningful, with 16 of 37 predicted pocket residues falling inside the annotated pocket.

Load-bearing premise

The load-bearing premise is that the residue-selection module actually learns from the interaction-prediction error, because the paper's masking step is a hard cutoff that blocks learning signals, and the paper does not describe a workaround; if no learning signal gets through, the module only copies conservation scores and the claimed benefit over a simple conservation filter disappears.

Editorial extensions

If this is right

  • If MIN's reported numbers hold, conservation-based residue screening gives a practical way to shorten long protein inputs and reduce noise, with the largest single-component AUC gain in the ablations.
  • A 0.5% ROC enrichment of 197.741 on DUD-E, if reproducible, means true actives would be concentrated near the top of a large decoy set, the operating regime that matters for virtual screening.
  • The multi-channel fusion claim implies sequence-only and structure-only representations carry complementary signals, so future DTI models can gain from combining them rather than choosing one input modality.
  • The case-study overlap suggests the same learned representations can double as a binding-pocket localiser without a separate pocket-detection stage.
  • The ablation ordering, where removing the C-Score Predictor drops AUC from 0.983 to 0.968, implies that the conservation-distillation module is the most influential component in the architecture.

Reading between the lines

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

  • The paper leaves implicit that Eq. (2) is a hard threshold, so the interaction-prediction loss cannot change the residue selector through it; a consequence is that, as stated, the predictor is trained only by the conservation regression, and the claimed joint optimization would require an explicit gradient bypass.
  • The paper does not say whether the contact maps use experimental or predicted protein structures; a testable extension is to compare the structure-aware channel's contribution under both structure sources.
  • The binding-pocket overlap is reported for a single structure; an extension is to run the same logit-difference procedure across many PDBbind complexes and report mean overlap, which would separate a generalisable behaviour from a favourable case.
  • The DUD-E evaluation splits targets into folds by protein family; an inference is that a split on unseen target families would test whether the reported AUC and enrichment transfer to genuinely novel proteins.
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

3 major / 6 minor

Summary. The paper proposes MIN, a deep-learning framework for drug-target interaction (DTI) classification. MIN has three main components: a representation-learning module that uses a C-Score Predictor to distill protein sequences by masking residues predicted to have low conservation scores, a multi-channel interaction module with structure-agnostic, structure-aware, and extended-mixture channels, and a contrastive-learning objective that aligns sequence and structure representations. Experiments on the DUD-E and Human benchmarks report state-of-the-art AUC, ROC enrichment, precision, and F1 scores, and a case study claims that the selected residues overlap known binding pockets. The central claims are that MIN outperforms all compared baselines and that the C-Score Predictor is trained jointly with the DTI classification objective while also using biological conservation information.

Significance. If the results are correct, MIN would be a strong DTI classifier, particularly because the DUD-E evaluation uses target-fold cross-validation, which is a more realistic and stringent setting than random splits. The multi-channel interaction design and the use of contrastive learning for representation alignment are reasonable and potentially useful ideas. The paper also reports an ablation study that quantifies the contribution of each component, which is valuable. However, the central novelty regarding the C-Score Predictor--that it is trained with the DTI objective to select residues for the DTI task--is not supported by the equations as written, and this directly affects the interpretability claim and the interpretation of the distillation mechanism.

major comments (3)
  1. [Section IV-A, Eq. (2), Eq. (10), Eq. (11), and Figure 4] The paper claims that the C-Score Predictor is 'also trained with the DTI objective' and 'influenced by DTI prediction objective.' However, Eq. (2) applies a hard threshold to the predicted C-Score: residues with predicted score above r0 are replaced by a fixed e[MASK] embedding. This operation is piecewise constant in the predicted score, so the derivative of the masked embedding with respect to the C-Score Predictor parameters is zero almost everywhere. Consequently, the cross-entropy loss LCE in Eq. (10) cannot provide any gradient to the C-Score Predictor through Eq. (2). The only training signal reaching the predictor is the MSE loss in Eq. (1), which fits ground-truth conservation scores. Unless a straight-through estimator, a Gumbel-Sigmoid relaxation, or another differentiable approximation is used--none of which is described--the 'Distill by C-Score Predictor' strategy in Figure 4 is simply a learned regressor of conservation scores, and the stated advantage over 'Distill by C-Score' in Section V-D is not explained. Please either describe a concrete differentiable mechanism that allows DTI gradients to reach the C-Score Predictor or revise the claims and the interpretation of the ablation and case study accordingly.
  2. [Section III-A(e) and Section V-A] The contact map P_e is built from distances between C-alpha atoms, but the paper does not state where the three-dimensional structures come from. For the DUD-E and Human datasets, are these experimental structures (e.g., from the PDB) or predicted structures (e.g., AlphaFold)? This matters for reproducibility, for the method's applicability to proteins without known structures, and for whether the comparison to baselines that do not use structures is fair. Please specify the structure source and, if predicted structures are used, the prediction pipeline and confidence criteria.
  3. [Section V-D, Case Study Q2] The case study numbers are internally inconsistent: the text says 'Among 37 residues in the binding pocket, 16 overlap residues' and then 'There are 37 residues in the real binding pocket and among them 9 residues are real binding positions.' If the predicted pocket contains 37 residues and the true pocket contains 37 residues, then '16 overlap' is the intersection size; but the next sentence says the real binding pocket has 9 residues. Please clarify the sizes of the real pocket, the predicted pocket, and the overlap, and also specify how the 'prediction binding residues' in Figure 6 are selected. Without this clarification, the explainability claim is hard to evaluate.
minor comments (6)
  1. [Section IV-A, Eq. (2)] The text says the model masks 'residues with low C-Score,' but Eq. (2) replaces residues with predicted C-Score greater than r0 with e[MASK] and keeps residues with score below or equal to r0. Since a lower C-Score is described as more conserved, the equation keeps conserved residues and masks highly variable ones. Please reword the sentence to avoid confusion.
  2. [Section V-A, Implementation Details] The contact threshold s0 in the definition of the contact map is never given a value. Only the conservation threshold r0 = 0.01 is reported. Please report s0 or state that it is a hyperparameter with its tuned value.
  3. [Section IV-C, Eq. (5)] The contrastive loss is defined with the symbol L used both for the denominator sum and for the loss function itself, making the expression difficult to parse. Please rename the denominator (e.g., S_neg) and make the positive/negative terms unambiguous.
  4. [Section V-B, Tables I and II] The dagger (†) indicates statistical significance with p < 0.01, but the statistical test is not described. Please specify the test (e.g., paired t-test across folds or seeds) and whether it is one-sided or two-sided.
  5. [Section V-A, Human dataset] The Human dataset is evaluated with a random split, which can overestimate performance because the same target may appear in both training and test sets. The authors follow previous work, so this is not a flaw by itself, but it should be acknowledged as a limitation with respect to generalization to unseen targets.
  6. [Section IV-D] There is a typo: 'tree-channel interaction network' should be 'three-channel interaction network.'

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the DTI prediction and the residue-selection component are not definitionally tied to their inputs; the hard-threshold gradient issue is a correctness gap, not a circular reduction.

full rationale

MIN's DTI output (Eq. 9) is a learned function of sequence and structure representations, and the DTI labels come from external benchmarks (DUD-E and Human) rather than from the model's own outputs. The C-Score labels are computed from external MSA/ConSurf analyses, so the residue-selection training signal in Eq. (1) is not derived from the DTI label. The case study uses a model trained on DUD-E and evaluates on PDBBind, which breaks any input/output equivalence. No load-bearing self-citation, imported uniqueness theorem, or ansatz-smuggling citation appears; the only self-citation overlap is a minor, non-load-bearing citation of [42] for the standard [CLS] token. One legitimate concern is that Eq. (2)'s hard threshold blocks direct gradients from Eq. (10) to the C-Score Predictor's own parameters phi_c and f_c, so the claim that the predictor is 'also trained with the DTI objective' (Section IV-A) is not supported by the written equations. This is an implementation/correctness issue, not a circularity, because the final DTI prediction is still an independent function of held-out labels and the residue selector is not used to construct those labels.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central empirical results rest on hand-chosen thresholds (r0, s0, tau), architecture choices, and three domain assumptions: conservation correlates with binding, contact maps capture structure, and reported baselines are comparable. No code or data is available to check these assumptions. The invented-entity count is zero because the C-Score Predictor and multi-channel interaction module are model components, not new physical or conceptual entities in the sense used here.

free parameters (4)
  • C-Score masking threshold r0 = 0.01
    Used in Eq. (2) to decide which residues are masked; no sensitivity analysis is reported and its value is inconsistent with the 588/1,608 residues retained in the case study.
  • InfoNCE temperature tau = not reported
    Temperature in contrastive losses Eq. (5); a standard default is not specified, so the contribution of contrastive learning is not reproducible.
  • Contact map threshold s0 = not reported
    Contact definition in Section III-A(e) leaves s0 unspecified, but the structure-aware channel depends on the resulting adjacency matrix.
  • Architecture hyperparameters = lr=1e-4, batch=256, embedding dim=64, target Transformer 2 layers/8 heads/hidden 256, drug Transformer 8 layers/8…
    Reported in Section V-A4; these hand-chosen values affect all reported results and are not ablated.
assumptions (5)
  • domain assumption Residues in binding pockets are more conserved than non-pocket residues, so conservation scores can guide residue selection.
    Section IV-A and Figure 3 support this statistically on PDBbind, but it is a tendency, not a guarantee, and the DTI objective is supposed to correct for exceptions.
  • domain assumption MSA-guided contact maps provide binding-relevant 3D structure information.
    Section III-A(e) defines contact maps from C-alpha distances, but the paper does not state the source of the 3D structures, experimental or predicted.
  • ad hoc to paper Hard masking in Eq. (2) permits end-to-end training of the C-Score Predictor with the DTI loss.
    Section IV-A and Eq. (11) assume gradients flow through the thresholded mask; no straight-through estimator or soft relaxation is described.
  • domain assumption Baseline results in Tables I and II are directly comparable despite coming from different papers and protocols.
    Section V-A1 describes the data splits but not a shared reimplementation or identical test-set negative sampling for all baselines.
  • domain assumption Splitting DUD-E folds by protein family removes homology leakage.
    Section V-A1; this is a reasonable standard practice but is not empirically validated in the paper.

how reviews work

0 comments
Cite this review

Pith. "Pith review of MIN: Multi-channel Interaction Network for Drug-Target Interaction with Protein Distillation." pith.science (2026). https://pith.science/paper/NQO4NTKW

@misc{pith2026241207778,
  author       = {Pith},
  title        = {Pith review of: MIN: Multi-channel Interaction Network for Drug-Target Interaction with Protein Distillation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NQO4NTKW}},
  note         = {Machine review of arXiv:2412.07778}
}
read the original abstract

Traditional drug discovery processes are both time-consuming and require extensive professional expertise. With the accumulation of drug-target interaction (DTI) data from experimental studies, leveraging modern machine-learning techniques to discern patterns between drugs and target proteins has become increasingly feasible. In this paper, we introduce the Multi-channel Interaction Network (MIN), a novel framework designed to predict DTIs through two primary components: a representation learning module and a multi-channel interaction module. The representation learning module features a C-Score Predictor-assisted screening mechanism, which selects critical residues to enhance prediction accuracy and reduce noise. The multi-channel interaction module incorporates a structure-agnostic channel, a structure-aware channel, and an extended-mixture channel, facilitating the identification of interaction patterns at various levels for optimal complementarity. Additionally, contrastive learning is utilized to harmonize the representations of diverse data types. Our experimental evaluations on public datasets demonstrate that MIN surpasses other strong DTI prediction methods. Furthermore, the case study reveals a high overlap between the residues selected by the C-Score Predictor and those in actual binding pockets, underscoring MIN's explainability capability. These findings affirm that MIN is not only a potent tool for DTI prediction but also offers fresh insights into the prediction of protein binding sites.

Figures

Figures reproduced from arXiv: 2412.07778 by the authors.

Figure 1
Figure 1. (a) A sketch map of the drug-target interaction classification task. (b) [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. An overview of MIN; (a) Target protein inputs; (b) Drug inputs; (c) Target protein representation learning; (d) Drug representation learning; (e) [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. The conservation score (C-Score) difference between the residues [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Performance comparison between C-Score Predictor with other [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: The comparison of the real binding pocket and prediction binding [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

55 extracted references · 53 canonical work pages

  1. [1]

    Multidimensional chemical genetic analysis of diversity-oriented synthesis-derived deacety- lase inhibitors using cell-based assays,

    S. J. Haggarty, K. M. Koeller, J. C. Wong, R. A. Butcher, and S. L. Schreiber, “Multidimensional chemical genetic analysis of diversity-oriented synthesis-derived deacety- lase inhibitors using cell-based assays,” Chemistry & biology, vol. 10, no. 5, pp. 383–396, 2003

    work page 2003

  2. [2]

    Drug–target interaction predic- tion: databases, web servers and computational models,

    X. Chen, C. C. Yan, X. Zhang, X. Zhang, F. Dai, J. Yin, and Y . Zhang, “Drug–target interaction predic- tion: databases, web servers and computational models,” Briefings in bioinformatics , vol. 17, no. 4, pp. 696–712, 2016

    work page 2016

  3. [3]

    Relating protein phar- macology by ligand chemistry,

    M. J. Keiser, B. L. Roth, B. N. Armbruster, P. Ernsberger, J. J. Irwin, and B. K. Shoichet, “Relating protein phar- macology by ligand chemistry,” Nature biotechnology , vol. 25, no. 2, pp. 197–206, 2007

    work page 2007

  4. [4]

    Predicting new molecular targets for known drugs,

    M. J. Keiser, V . Setola, J. J. Irwin, C. Laggner, A. I. Abbas, S. J. Hufeisen, N. H. Jensen, M. B. Kuijer, R. C. Matos, T. B. Tran et al. , “Predicting new molecular targets for known drugs,” Nature, vol. 462, no. 7270, pp. 175–181, 2009

    work page 2009

  5. [5]

    Structure-based maximal affinity model pre- dicts small-molecule druggability,

    A. C. Cheng, R. G. Coleman, K. T. Smyth, Q. Cao, P. Soulard, D. R. Caffrey, A. C. Salzberg, and E. S. Huang, “Structure-based maximal affinity model pre- dicts small-molecule druggability,” Nature biotechnol- ogy, vol. 25, no. 1, pp. 71–75, 2007

    work page 2007

  6. [6]

    B. R. Donald, Algorithms in structural molecular biol- ogy. MIT Press, 2011

    work page 2011

  7. [7]

    Predictive power of different types of experimental restraints in small molecule dock- ing: A review,

    D. Y . Fu and J. Meiler, “Predictive power of different types of experimental restraints in small molecule dock- ing: A review,” Journal of chemical information and modeling, vol. 58, no. 2, pp. 225–233, 2018

    work page 2018

  8. [8]

    Recommen- dation techniques for drug–target interaction prediction and drug repositioning,

    S. Alaimo, R. Giugno, and A. Pulvirenti, “Recommen- dation techniques for drug–target interaction prediction and drug repositioning,” in Data Mining Techniques for the Life Sciences . Springer, 2016, pp. 441–462

    work page 2016

Show all 55 references
  1. [9]

    Deep-learning-based drug–target interaction prediction,

    M. Wen, Z. Zhang, S. Niu, H. Sha, R. Yang, Y . Yun, and H. Lu, “Deep-learning-based drug–target interaction prediction,” Journal of proteome research, vol. 16, no. 4, pp. 1401–1409, 2017

    2017

  2. [10]

    Supervised prediction of drug–target interactions using bipartite local models,

    K. Bleakley and Y . Yamanishi, “Supervised prediction of drug–target interactions using bipartite local models,” Bioinformatics, vol. 25, no. 18, pp. 2397–2403, 2009

    2009

  3. [11]

    Smiles, a chemical language and informa- tion system. 1. introduction to methodology and encoding rules,

    D. Weininger, “Smiles, a chemical language and informa- tion system. 1. introduction to methodology and encoding rules,” Journal of chemical information and computer sciences, vol. 28, no. 1, pp. 31–36, 1988

    1988

  4. [12]

    Protein–ligand scoring with convolutional neural networks,

    M. Ragoza, J. Hochuli, E. Idrobo, J. Sunseri, and D. R. Koes, “Protein–ligand scoring with convolutional neural networks,” Journal of chemical information and model- ing, vol. 57, no. 4, pp. 942–957, 2017

    2017

  5. [13]

    Development and evaluation of a deep learning model for protein–ligand binding affinity pre- diction,

    M. M. Stepniewska-Dziubinska, P. Zielenkiewicz, and P. Siedlecki, “Development and evaluation of a deep learning model for protein–ligand binding affinity pre- diction,” Bioinformatics, vol. 34, no. 21, pp. 3666–3674, 2018

    2018

  6. [14]

    Graph convolutional neural networks for predicting drug-target interactions,

    W. Torng and R. B. Altman, “Graph convolutional neural networks for predicting drug-target interactions,” Journal of chemical information and modeling , vol. 59, no. 10, pp. 4131–4149, 2019

    2019

  7. [15]

    Predict- ing drug–protein interaction using quasi-visual question answering system,

    S. Zheng, Y . Li, S. Chen, J. Xu, and Y . Yang, “Predict- ing drug–protein interaction using quasi-visual question answering system,” Nature Machine Intelligence , vol. 2, no. 2, pp. 134–140, 2020

    2020

  8. [16]

    Deepdta: deep drug–target binding affinity prediction,

    H. ¨Ozt¨urk, A. ¨Ozg¨ur, and E. Ozkirimli, “Deepdta: deep drug–target binding affinity prediction,” Bioinformatics, vol. 34, no. 17, pp. i821–i829, 2018

    2018

  9. [17]

    Deepaffinity: interpretable deep learning of compound–protein affinity through unified recurrent and convolutional neural net- works,

    M. Karimi, D. Wu, Z. Wang, and Y . Shen, “Deepaffinity: interpretable deep learning of compound–protein affinity through unified recurrent and convolutional neural net- works,” Bioinformatics, vol. 35, no. 18, pp. 3329–3338, 2019

    2019

  10. [18]

    Deeppurpose: a deep learning library for drug-target interaction prediction,

    K. Huang, T. Fu, L. Glass, M. Zitnik, C. Xiao, and J. Sun, “Deeppurpose: a deep learning library for drug-target interaction prediction,” Bioinformatics, 2020

    2020

  11. [19]

    Monn: a multi-objective neural network for predicting compound-protein interactions and affinities,

    S. Li, F. Wan, H. Shu, T. Jiang, D. Zhao, and J. Zeng, “Monn: a multi-objective neural network for predicting compound-protein interactions and affinities,” Cell Sys- tems, vol. 10, no. 4, pp. 308–322, 2020

    2020

  12. [20]

    Semi-supervised classifica- tion with graph convolutional networks,

    T. N. Kipf and M. Welling, “Semi-supervised classifica- tion with graph convolutional networks,” arXiv preprint arXiv:1609.02907, 2016

    2016 arXiv

  13. [21]

    Nnscore 2.0: A neural-network receptor–ligand scoring function,

    J. D. Durrant and J. A. Mccammon, “Nnscore 2.0: A neural-network receptor–ligand scoring function,” Jour- nal of Chemical Information & Modeling, vol. 51, no. 11, p. 2897, 2011

    2011

  14. [22]

    Msa-regularized protein sequence transformer toward predicting genome-wide chemical-protein interac- tions: Application to gpcrome deorphanization,

    T. Cai, H. Lim, K. A. Abbu, Y . Qiu, R. Nussinov, and L. Xie, “Msa-regularized protein sequence transformer toward predicting genome-wide chemical-protein interac- tions: Application to gpcrome deorphanization,” Journal of chemical information and modeling, vol. 61, no. 4, pp....

    2021

  15. [23]

    Transformer- cpi: improving compound–protein interaction predic- tion by sequence-based deep learning with self-attention mechanism and label reversal experiments,

    L. Chen, X. Tan, D. Wang, F. Zhong, X. Liu, T. Yang, X. Luo, K. Chen, H. Jiang, and M. Zheng, “Transformer- cpi: improving compound–protein interaction predic- tion by sequence-based deep learning with self-attention mechanism and label reversal experiments,” Bioinformat- ics,...

    2020

  16. [24]

    Gandti: A multi-task neural network for drug- target interaction prediction,

    S. Wang, P. Shan, Y . Zhao, and L. Zuo, “Gandti: A multi-task neural network for drug- target interaction prediction,” Computational Biology and Chemistry , vol. 92, p. 107476, 2021. [Online]. Available: https://www.sciencedirect.com/science/article/ pii/S1476927121000438

    2021

  17. [25]

    Bridgedpi: a novel graph neural network for predicting drug–protein interactions,

    Y . Wu, M. Gao, M. Zeng, J. Zhang, and M. Li, “Bridgedpi: a novel graph neural network for predicting drug–protein interactions,” Bioinformatics, vol. 38, no. 9, pp. 2571–2578, 2022

    2022

  18. [26]

    Attentionsit- edti: an interpretable graph-based model for drug-target interaction prediction using nlp sentence-level relation classification,

    M. Yazdani-Jahromi, N. Yousefi, A. Tayebi, E. Kolanthai, C. J. Neal, S. Seal, and O. O. Garibay, “Attentionsit- edti: an interpretable graph-based model for drug-target interaction prediction using nlp sentence-level relation classification,” Briefings in Bioinformatics , 2022

    2022

  19. [27]

    Pignet: a physics-informed deep learning model toward generalized drug–target interaction predictions,

    S. Moon, W. Zhung, S. Yang, J. Lim, and W. Y . Kim, “Pignet: a physics-informed deep learning model toward generalized drug–target interaction predictions,” Chemi- JOURNAL OF LATEX CLASS FILES, VOL. 18, NO. 9, SEPTEMBER 2020 10 cal Science, vol. 13, no. 13, pp. 3661–3673, 2022

    2020

  20. [28]

    Deepfusion: A deep learning based multi-scale feature fusion method for predicting drug-target interactions,

    T. Song, X. Zhang, M. Ding, A. Rodriguez- Paton, S. Wang, and G. Wang, “Deepfusion: A deep learning based multi-scale feature fusion method for predicting drug-target interactions,” Methods, vol. 204, pp. 269–277, 2022. [Online]. Available: https://www.sciencedirect.com/scienc...

    2022

  21. [29]

    Tripletmultidti: multimodal representation learning in drug-target interaction prediction with triplet loss func- tion,

    A. Dehghan, P. Razzaghi, K. Abbasi, and S. Gharaghani, “Tripletmultidti: multimodal representation learning in drug-target interaction prediction with triplet loss func- tion,” Expert Systems with Applications , vol. 232, p. 120754, 2023

    2023

  22. [30]

    Gcardti: Drug–target interaction prediction based on a hybrid mechanism in drug selfies,

    Y . Feng, Y . Zhang, Z. Deng, and M. Xiong, “Gcardti: Drug–target interaction prediction based on a hybrid mechanism in drug selfies,” Quantitative Biology, 2024

    2024

  23. [31]

    Hi- graphdti: Hierarchical graph representation learning for drug-target interaction prediction,

    B. Liu, S. Wu, J. Wang, X. Deng, and A. Zhou, “Hi- graphdti: Hierarchical graph representation learning for drug-target interaction prediction,” in Joint European Conference on Machine Learning and Knowledge Dis- covery in Databases . Springer, 2024, pp. 354–370

    2024

  24. [32]

    Pocketdta: an advanced multimodal architecture for enhanced prediction of drug- target affinity from 3d structural data of target binding pockets,

    L. Zhao, H. Wang, and S. Shi, “Pocketdta: an advanced multimodal architecture for enhanced prediction of drug- target affinity from 3d structural data of target binding pockets,” Bioinformatics, vol. 40, no. 10, p. btae594, 2024

    2024

  25. [33]

    Drugclip: Contrasive protein- molecule representation learning for virtual screening,

    B. Gao, B. Qiang, H. Tan, Y . Jia, M. Ren, M. Lu, J. Liu, W.-Y . Ma, and Y . Lan, “Drugclip: Contrasive protein- molecule representation learning for virtual screening,” Advances in Neural Information Processing Systems , vol. 36, 2024

    2024

  26. [34]

    Hyperpcm: Robust task-conditioned modeling of drug–target interactions,

    E. Svensson, P.-J. Hoedt, S. Hochreiter, and G. Klam- bauer, “Hyperpcm: Robust task-conditioned modeling of drug–target interactions,” Journal of Chemical Informa- tion and Modeling , vol. 64, no. 7, pp. 2539–2553, 2024

    2024

  27. [35]

    Uniclust databases of clustered and deeply annotated protein sequences and alignments,

    M. Mirdita, L. von den Driesch, C. Galiez, M. J. Martin, J. S ¨oding, and M. Steinegger, “Uniclust databases of clustered and deeply annotated protein sequences and alignments,” Nucleic Acids Research , vol. 45, no. D1, pp. D170–D176, 11 2016. [Online]. Available: https://doi....

    2016 doi

  28. [36]

    Predicting functionally important residues from sequence conservation,

    J. A. Capra and M. Singh, “Predicting functionally important residues from sequence conservation,” Bioin- formatics, vol. 23, no. 15, pp. 1875–1882, 05 2007

    2007

  29. [37]

    Attention is all you need,

    A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomez, Ł. Kaiser, and I. Polosukhin, “Attention is all you need,” in Advances in neural infor- mation processing systems , 2017, pp. 5998–6008

    2017

  30. [38]

    Conservation and rela- tive importance of residues across protein-protein inter- faces,

    M. Guharoy and P. Chakrabarti, “Conservation and rela- tive importance of residues across protein-protein inter- faces,” Proceedings of the National Academy of Sciences, vol. 102, no. 43, pp. 15 447–15 452, 2005

    2005

  31. [39]

    The pdbbind database: Collection of binding affinities for protein- ligand complexes with known three-dimensional struc- tures,

    R. Wang, X. Fang, Y . Lu, and S. Wang, “The pdbbind database: Collection of binding affinities for protein- ligand complexes with known three-dimensional struc- tures,” Journal of medicinal chemistry , vol. 47, no. 12, pp. 2977–2980, 2004

    2004

  32. [40]

    Consurf-db: An accessible repository for the evolution- ary conservation patterns of the majority of pdb proteins,

    A. Ben Chorin, G. Masrati, A. Kessel, A. Narunsky, J. Sprinzak, S. Lahav, H. Ashkenazy, and N. Ben-Tal, “Consurf-db: An accessible repository for the evolution- ary conservation patterns of the majority of pdb proteins,” Protein Science, vol. 29, no. 1, pp. 258–267, 2020

    2020

  33. [41]

    The consurf-db: pre-calculated evolutionary conserva- tion profiles of protein structures,

    O. Goldenberg, E. Erez, G. Nimrod, and N. Ben-Tal, “The consurf-db: pre-calculated evolutionary conserva- tion profiles of protein structures,” Nucleic acids re- search, vol. 37, no. suppl 1, pp. D323–D327, 2009

    2009

  34. [42]

    Dual-view molecule pre-training,

    J. Zhu, Y . Xia, T. Qin, W. Zhou, H. Li, and T.-Y . Liu, “Dual-view molecule pre-training,” arXiv preprint arXiv:2106.10234, 2021

    2021 arXiv

  35. [43]

    Adaptive graph contrastive learning for recommendation,

    Y . Jiang, C. Huang, and L. Huang, “Adaptive graph contrastive learning for recommendation,” inProceedings of the 29th ACM SIGKDD conference on knowledge discovery and data mining , 2023, pp. 4252–4261

    2023

  36. [44]

    Simple and asymmetric graph contrastive learning without augmen- tations,

    T. Xiao, H. Zhu, Z. Chen, and S. Wang, “Simple and asymmetric graph contrastive learning without augmen- tations,” Advances in Neural Information Processing Systems, vol. 36, 2024

    2024

  37. [45]

    Mo- mentum contrast for unsupervised visual representation learning,

    K. He, H. Fan, Y . Wu, S. Xie, and R. Girshick, “Mo- mentum contrast for unsupervised visual representation learning,” in Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition , 2020, pp. 9729–9738

    2020

  38. [46]

    Improv- ing compound–protein interaction prediction by building up highly credible negative samples,

    H. Liu, J. Sun, J. Guan, J. Zheng, and S. Zhou, “Improv- ing compound–protein interaction prediction by building up highly credible negative samples,” Bioinformatics, vol. 31, no. 12, pp. i221–i229, 2015

    2015

  39. [47]

    Compound–protein interaction prediction with end-to-end learning of neu- ral networks for graphs and sequences,

    M. Tsubaki, K. Tomii, and J. Sese, “Compound–protein interaction prediction with end-to-end learning of neu- ral networks for graphs and sequences,” Bioinformatics, vol. 35, no. 2, pp. 309–318, 2019

    2019

  40. [48]

    Autodock vina: improving the speed and accuracy of docking with a new scoring func- tion, efficient optimization, and multithreading,

    O. Trott and A. J. Olson, “Autodock vina: improving the speed and accuracy of docking with a new scoring func- tion, efficient optimization, and multithreading,” Journal of computational chemistry , vol. 31, no. 2, pp. 455–461, 2010

    2010

  41. [49]

    Nnscore 2.0: a neural-network receptor–ligand scoring function,

    J. D. Durrant and J. A. McCammon, “Nnscore 2.0: a neural-network receptor–ligand scoring function,” Jour- nal of chemical information and modeling , vol. 51, no. 11, pp. 2897–2903, 2011

    2011

  42. [50]

    A machine learning approach to predicting protein–ligand binding affinity with applications to molecular docking,

    P. J. Ballester and J. B. Mitchell, “A machine learning approach to predicting protein–ligand binding affinity with applications to molecular docking,” Bioinformatics, vol. 26, no. 9, pp. 1169–1175, 2010

    2010

  43. [51]

    The weka data mining soft- ware: an update,

    M. Hall, E. Frank, G. Holmes, B. Pfahringer, P. Reute- mann, and I. H. Witten, “The weka data mining soft- ware: an update,” ACM SIGKDD explorations newsletter, vol. 11, no. 1, pp. 10–18, 2009

    2009

  44. [52]

    Liblinear: A library for large linear classifica- tion,

    R.-E. Fan, K.-W. Chang, C.-J. Hsieh, X.-R. Wang, and C.-J. Lin, “Liblinear: A library for large linear classifica- tion,” the Journal of machine Learning research , vol. 9, pp. 1871–1874, 2008

    2008

  45. [53]

    Graphdta: predic- tion of drug–target binding affinity using graph convolu- tional networks,

    T. Nguyen, H. Le, and S. Venkatesh, “Graphdta: predic- tion of drug–target binding affinity using graph convolu- tional networks,” BioRxiv, p. 684662, 2019

    2019

  46. [54]

    Adam: A method for stochastic optimization,

    D. P. Kingma and J. Ba, “Adam: A method for stochastic optimization,” arXiv preprint arXiv:1412.6980 , 2014. JOURNAL OF LATEX CLASS FILES, VOL. 18, NO. 9, SEPTEMBER 2020 11

    2014 arXiv

  47. [55]

    The pdbbind database: methodologies and updates,

    R. Wang, X. Fang, Y . Lu, C.-Y . Yang, and S. Wang, “The pdbbind database: methodologies and updates,” Journal of medicinal chemistry , vol. 48, no. 12, pp. 4111–4119, 2005. Shuqi Li is currently working toward a Ph.D. degree with the Gaoling School of Artificial Intelligence ...

    2005

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.