Drug–Target Interaction Prediction Based on an Interactive Inference Network
"> Figure 1
<p>ROC curves of INDTI with different combinations of input and encoder and comparison to other algorithms. The solid lines indicate different algorithms, and the dashed line indicates random selection. The values in parentheses represent the area under the curve (AUC), which is a metric used to evaluate the performance of classifiers. The different combinations of modules used in INDTI are also listed in parentheses.</p> "> Figure 2
<p>PR-AUC of different protein families. The solid line indicates different protein families. The values in parentheses represent the area under the precision–recall (PR) curve, which is a typical way to summarize a model’s overall performance.</p> "> Figure 3
<p>ROC-AUC of different protein families. The solid line indicates different protein families, and the dashed line indicates random selection. The values in parentheses represent the area under the curve (AUC).</p> "> Figure 4
<p>Drug–target interaction space diagram. (<b>a</b>) Pidolic acid interacts with Orexin. (<b>b</b>) 4-hydroxybenzaldehyde O-(3,3-dimethylbutanoyl)OXIME interacts with macrophage migration inhibitor factor. (<b>c</b>) 5-benzyl-1,3-thiazol-2-amine interacts with camp dependent protein kinase inhibitors. (<b>d</b>) 4-(4-chlorobenzyl)-1-(7H-Pyrrolo[2,3-D]Pyrimidin-4-yl)piperidin-4-aminium interacts with camp dependent protein kinase inhibitors. The different colors in the key represent the intensity of the absolute value of interaction value which was extracted from the hidden representation of the interaction layer.</p> "> Figure 5
<p>Prediction of Alzheimer’s disease-related drug–target interactions. The red dots in the interaction network diagram represent the 22 Alzheimer’s-related targets in <a href="#ijms-25-07753-t004" class="html-table">Table 4</a>. Each light purple square represents a drug, and the edge represents an interaction between the drug and the target. The edge emitted by each target is given a different color.</p> "> Figure 6
<p>Venn diagram of relationships of DAVIS, BindingDB, and BIOSNAP datasets.</p> "> Figure 7
<p>Model architecture of INDTI.</p> "> Figure 8
<p>Molecular sub-sequence embedding process.</p> ">
Abstract
:1. Introduction
2. Results
2.1. Performance Comparison between INDTI and Other Models
- DeepDTI [25] incorporated deep belief networks (DBNs) for DTI prediction. The combination of ECFP2, ECFP4, and ECFP6 was used as the drug input, and a protein sequence composition (PSC) was used as the target input. A PSC is composed of an amino acid composition (AAC), a dipeptide composition (DC), and a tripeptide composition (TC), which are, respectively, the frequencies of one, two, and three amino acids.
- DeepConv-DTI [19] used a CNN to extract local features of various lengths in amino acid sequences of targets and ECFP4 fingerprints of drugs and then input the combined feature vectors into the full connectivity layer to achieve prediction results.
- DeepDTA [18] employed a CNN to analyze SMILES strings and target amino acid sequences. In order to obtain the prediction results, the combined vectors were input to the full connection layer. A Sigmoid activation function was attached to its full connection layer for the DeepDTA task to be suitable for the binary prediction task in this publication.
2.2. Robustness
2.3. Interpretability
2.4. Prediction and Validation of Alzheimer’s Disease-Related Targets
3. Discussion
4. Materials and Methods
4.1. Datasets
4.2. Methods
4.2.1. Embedding Layer
4.2.2. Encoding Layer
4.2.3. Drug–Target Interaction Prediction
4.2.4. Input Representation
Supplementary Materials
Author Contributions
Funding
Institutional Review Board Statement
Informed Consent Statement
Data Availability Statement
Acknowledgments
Conflicts of Interest
References
- Allen, J.A.; Roth, B.L. Strategies to discover unexpected targets for drugs active at G protein–coupled receptors. Annu. Rev. Pharmacol. Toxicol. 2011, 51, 117–144. [Google Scholar] [CrossRef] [PubMed]
- Cheng, F.; Liu, C.; Jiang, J.; Lu, W.; Li, W.; Liu, G.; Zhou, W.; Huang, J.; Tang, Y. Prediction of drug-target interactions and drug repositioning via network-based inference. PLoS Comput. Biol. 2012, 8, e1002503. [Google Scholar] [CrossRef] [PubMed]
- Landry, Y.; Gies, J.P. Drugs and their molecular targets: An updated overview. Fundam. Clin. Pharmacol. 2008, 22, 1–18. [Google Scholar] [CrossRef] [PubMed]
- Hopkins, A.L.; Keserü, G.M.; Leeson, P.D.; Rees, D.C.; Reynolds, C.H. The role of ligand efficiency metrics in drug discovery. Nat. Rev. Drug Discov. 2014, 13, 105–121. [Google Scholar] [CrossRef] [PubMed]
- Alonso, H.; Bliznyuk, A.A.; Gready, J.E. Combining docking and molecular dynamic simulations in drug design. Med. Res. Rev. 2006, 26, 531–568. [Google Scholar] [CrossRef] [PubMed]
- Ezzat, A.; Wu, M.; Li, X.L.; Kwoh, C.K. Computational prediction of drug-target interactions using chemogenomic approaches: An empirical survey. Brief. Bioinform. 2019, 20, 1337–1357. [Google Scholar] [CrossRef] [PubMed]
- Bagherian, M.; Sabeti, E.; Wang, K.; Sartor, M.A.; Nikolovska-Coleska, Z.; Najarian, K. Machine learning approaches and databases for prediction of drug-target interaction: A survey paper. Brief. Bioinform. 2021, 22, 247–269. [Google Scholar] [CrossRef] [PubMed]
- Crampon, K.; Giorkallos, A.; Deldossi, M.; Baud, S.; Steffenel, L.A. Machine-learning methods for ligand-protein molecular docking. Drug Discov. Today 2022, 27, 151–164. [Google Scholar] [CrossRef] [PubMed]
- Yazdani-Jahromi, M.; Yousefi, N.; Tayebi, A.; Kolanthai, E.; Neal, C.J.; Seal, S.; Garibay, O.O. AttentionSiteDTI: An interpretable graph-based model for drug-target interaction prediction using NLP sentence-level relation classification. Brief. Bioinform. 2022, 23, bbac272. [Google Scholar] [CrossRef]
- Smith, Z.; Strobel, M.; Vani, B.P.; Tiwary, P. Graph Attention Site Prediction (GrASP): Identifying Druggable Binding Sites Using Graph Neural Networks with Attention. bioRxiv 2023. [Google Scholar] [CrossRef]
- Tang, X.; Lei, X.; Zhang, Y. Prediction of Drug-Target Affinity Using Attention Neural Network. Int. J. Mol. Sci. 2024, 25, 5126. [Google Scholar] [CrossRef] [PubMed]
- Nath, A.; Kumari, P.; Chaube, R. Prediction of human drug targets and their interactions using machine learning methods: Current and future perspectives. In Computational Drug Discovery and Design; Humana: New York, NY, USA, 2018; pp. 21–30. [Google Scholar]
- Yamanishi, Y. Chemogenomic approaches to infer drug–target interaction networks. In Data Mining for Systems Biology: Methods and Protocols; Humana: Totowa, NJ, USA, 2013; pp. 97–113. [Google Scholar]
- Mousavian, Z.; Masoudi-Nejad, A. Drug–target interaction prediction via chemogenomic space: Learning-based methods. Expert Opin. Drug Metab. Toxicol. 2014, 10, 1273–1287. [Google Scholar] [CrossRef] [PubMed]
- Nagamine, N.; Sakakibara, Y. Statistical prediction of protein–chemical interactions based on chemical structure and mass spectrometry data. Bioinformatics 2007, 23, 2004–2012. [Google Scholar] [CrossRef] [PubMed]
- He, T.; Heidemeyer, M.; Ban, F.; Cherkasov, A.; Ester, M. SimBoost: A read-across approach for predicting drug–target binding affinities using gradient boosting machines. J. Cheminform. 2017, 9, 24. [Google Scholar] [CrossRef] [PubMed]
- Luo, Y.; Zhao, X.; Zhou, J.; Yang, J.; Zhang, Y.; Kuang, W.; Peng, J.; Chen, L.; Zeng, J. A network integration approach for drug-target interaction prediction and computational drug repositioning from heterogeneous information. Nat. Commun. 2017, 8, 573. [Google Scholar] [CrossRef] [PubMed]
- Öztürk, H.; Özgür, A.; Ozkirimli, E. DeepDTA: Deep drug–target binding affinity prediction. Bioinformatics 2018, 34, i821–i829. [Google Scholar] [CrossRef] [PubMed]
- Lee, I.; Keum, J.; Nam, H. DeepConv-DTI: Prediction of drug-target interactions via deep learning with convolution on protein sequences. PLoS Comput. Biol. 2019, 15, e1007129. [Google Scholar] [CrossRef] [PubMed]
- Nguyen, T.; Le, H.; Quinn, T.P.; Nguyen, T.; Le, T.D.; Venkatesh, S. GraphDTA: Predicting drug–target binding affinity with graph neural networks. Bioinformatics 2021, 37, 1140–1147. [Google Scholar] [CrossRef] [PubMed]
- Moon, S.; Zhung, W.; Yang, S.; Lim, J.; Kim, W.Y. PIGNet: A physics-informed deep learning model toward generalized drug–target interaction predictions. Chem. Sci. 2022, 13, 3661–3673. [Google Scholar] [CrossRef]
- Zhao, B.-W.; Su, X.-R.; Hu, P.-W.; Huang, Y.-A.; You, Z.-H.; Hu, L. iGRLDTI: An improved graph representation learning method for predicting drug–target interactions over heterogeneous biological information network. Bioinformatics 2023, 39, btad451. [Google Scholar] [CrossRef]
- Gong, Y.; Luo, H.; Zhang, J. Natural language inference over interaction space. arXiv 2017, arXiv:1709.04348. [Google Scholar]
- Huang, K.; Xiao, C.; Glass, L.; Sun, J. Explainable Substructure Partition Fingerprint for Protein, Drug, and More. In Proceedings of the NeurIPS Learning Meaningful Representation of Life Workshop, Vancouver, BC, Canada, 13 December 2019. [Google Scholar]
- Wen, M.; Zhang, Z.; Niu, S.; Sha, H.; Yang, R.; Yun, Y.; Lu, H. Deep-learning-based drug–target interaction prediction. J. Proteome Res. 2017, 16, 1401–1409. [Google Scholar] [CrossRef] [PubMed]
- Xie, J.; Liang, R.; Wang, Y.; Huang, J.; Cao, X.; Niu, B. Progress in target drug molecules for Alzheimer’s disease. Curr. Top. Med. Chem. 2020, 20, 4–36. [Google Scholar] [CrossRef] [PubMed]
- Quartey, M.O.; Nyarko, J.N.; Maley, J.M.; Barnes, J.R.; Bolanos, M.A.; Heistad, R.M.; Knudsen, K.J.; Pennington, P.R.; Buttigieg, J.; De Carvalho, C.E. The Aβ (1–38) peptide is a negative regulator of the Aβ (1–42) peptide implicated in Alzheimer disease progression. Sci. Rep. 2021, 11, 431. [Google Scholar] [CrossRef] [PubMed]
- Urban, A.S.; Pavlov, K.V.; Kamynina, A.V.; Okhrimenko, I.S.; Arseniev, A.S.; Bocharov, E.V. Structural studies providing insights into production and conformational behavior of amyloid-β peptide associated with Alzheimer’s disease development. Molecules 2021, 26, 2897. [Google Scholar] [CrossRef] [PubMed]
- Muralidar, S.; Ambi, S.V.; Sekaran, S.; Thirumalai, D.; Palaniappan, B. Role of tau protein in Alzheimer’s disease: The prime pathological player. Int. J. Biol. Macromol. 2020, 163, 1599–1617. [Google Scholar] [CrossRef] [PubMed]
- Pîrşcoveanu, D.F.V.; Pirici, I.; Tudorică, V.; Bălşeanu, T.-A.; Albu, V.-C.; Bondari, S.; Bumbea, A.M.; Pîrşcoveanu, M. Tau protein in neurodegenerative diseases—A review. Rom. J. Morphol. Embryol. 2017, 58, 1141–1150. [Google Scholar] [PubMed]
- Roda, A.R.; Serra-Mir, G.; Montoliu-Gaya, L.; Tiessler, L.; Villegas, S. Amyloid-beta peptide and tau protein crosstalk in Alzheimer’s disease. Neural Regen. Res. 2022, 17, 1666–1674. [Google Scholar] [PubMed]
- Tang, M.-X.; Stern, Y.; Marder, K.; Bell, K.; Gurland, B.; Lantigua, R.; Andrews, H.; Feng, L.; Tycko, B.; Mayeux, R. The APOE-∊ 4 allele and the risk of Alzheimer disease among African Americans, whites, and Hispanics. JAMA 1998, 279, 751–755. [Google Scholar] [CrossRef]
- Raulin, A.-C.; Doss, S.V.; Trottier, Z.A.; Ikezu, T.C.; Bu, G.; Liu, C.-C. ApoE in Alzheimer’s disease: Pathophysiology and therapeutic strategies. Mol. Neurodegener. 2022, 17, 72. [Google Scholar] [CrossRef]
- Zitnik, M.; Sosic, R.; Leskovec, J. BioSNAP Datasets: Stanford Biomedical Network Dataset Collection. 2018. Available online: http://snap.stanford.edu/biodata (accessed on 8 October 2022).
- Gilson, M.K.; Liu, T.; Baitaluk, M.; Nicola, G.; Hwang, L.; Chong, J. BindingDB in 2015: A public database for medicinal chemistry, computational chemistry and systems pharmacology. Nucleic Acids Res. 2016, 44, D1045–D1053. [Google Scholar] [PubMed]
- Davis, M.I.; Hunt, J.P.; Herrgard, S.; Ciceri, P.; Wodicka, L.M.; Pallares, G.; Hocker, M.; Treiber, D.K.; Zarrinkar, P.P. Comprehensive analysis of kinase inhibitor selectivity. Nat. Biotechnol. 2011, 29, 1046–1051. [Google Scholar] [CrossRef] [PubMed]
- Wishart, D.S.; Knox, C.; Guo, A.C.; Cheng, D.; Shrivastava, S.; Tzur, D.; Gautam, B.; Hassanali, M. DrugBank: A knowledgebase for drugs, drug actions and drug targets. Nucleic Acids Res. 2008, 36, D901–D906. [Google Scholar] [CrossRef] [PubMed]
- Boutet, E.; Lieberherr, D.; Tognolli, M.; Schneider, M.; Bairoch, A. UniProtKB/Swiss-Prot: The manually annotated section of the UniProt KnowledgeBase. In Plant Bioinformatics: Methods and Protocols; Springer: Berlin/Heidelberg, Germany, 2007; pp. 89–112. [Google Scholar]
- Gaulton, A.; Bellis, L.J.; Bento, A.P.; Chambers, J.; Davies, M.; Hersey, A.; Light, Y.; McGlinchey, S.; Michalovich, D.; Al-Lazikani, B. ChEMBL: A large-scale bioactivity database for drug discovery. Nucleic Acids Res. 2012, 40, D1100–D1107. [Google Scholar] [CrossRef] [PubMed]
- Gao, K.Y.; Fokoue, A.; Luo, H.; Iyengar, A.; Dey, S.; Zhang, P. Interpretable drug target prediction using deep neural representation. In Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence, IJCAI, Stockholm, Sweden, 13–19 July 2018; pp. 3371–3377. [Google Scholar]
- Huang, K.; Fu, T.; Glass, L.M.; Zitnik, M.; Xiao, C.; Sun, J. DeepPurpose: A deep learning library for drug–target interaction prediction. Bioinformatics 2020, 36, 5545–5547. [Google Scholar] [CrossRef]
- Rogers, D.; Hahn, M. Extended-connectivity fingerprints. J. Chem. Inf. Model. 2010, 50, 742–754. [Google Scholar] [CrossRef]
- Bolton, E.E.; Wang, Y.; Thiessen, P.A.; Bryant, S.H. PubChem: Integrated platform of small molecules and biological activities. In Annual Reports in Computational Chemistry; Elsevier: Amsterdam, The Netherlands, 2008; Volume 4, pp. 217–241. [Google Scholar]
Drug Input | Target Input | Encoder |
---|---|---|
SMILES | Amino acid sequences | CNN |
SMILES + Morgan | CNN, MLP | |
SMILES + PubChem | CNN, MLP | |
SMILES | Self-Attention | |
SMILES + Morgan | ||
SMILES + PubChem |
Model | Accuracy | Precision | Recall | Specificity | F1 | MCC | |
---|---|---|---|---|---|---|---|
INDTI | CNN | 0.820 | 0.514 | 0.862 | 0.810 | 0.644 | 0.566 |
Morgan and CNN | 0.813 | 0.504 | 0.839 | 0.807 | 0.629 | 0.545 | |
PubChem and CNN | 0.828 | 0.528 | 0.874 | 0.818 | 0.658 | 0.584 | |
Self-Attention | 0.812 | 0.502 | 0.780 | 0.819 | 0.611 | 0.515 | |
Morgan and Self-Attention | 0.811 | 0.500 | 0.796 | 0.814 | 0.614 | 0.521 | |
PubChem and Self-Attention | 0.800 | 0.482 | 0.816 | 0.800 | 0.606 | 0.513 | |
DeepDTA | 0.810 | 0.512 | 0.853 | 0.806 | 0.630 | 0.564 | |
DeepConv-DTI | 0.789 | 0.469 | 0.834 | 0.779 | 0.600 | 0.508 | |
DeepDTI | 0.700 | 0.378 | 0.897 | 0.780 | 0.533 | 0.441 |
Protein Families | Accuracy | Precision | Recall | F1 | MCC |
---|---|---|---|---|---|
All | 0.828 | 0.528 | 0.874 | 0.658 | 0.584 |
Catalytic Receptor | 0.794 | 0.480 | 0.870 | 0.619 | 0.534 |
Enzyme | 0.820 | 0.588 | 0.880 | 0.704 | 0.606 |
GPCRs | 0.744 | 0.679 | 0.958 | 0.794 | 0.532 |
Ion Channel | 0.755 | 0.649 | 0.967 | 0.777 | 0.581 |
Nuclear Receptor | 0.695 | 0.600 | 0.928 | 0.727 | 0.470 |
Target Gene | Target |
---|---|
AChE | Acetylcholinesterase |
BChE | Butyrylcholinesterase |
App | Amyloid-β |
BACE1 | Aspartyl protease 2 |
SSTR4 | Somatostatin receptor-4 |
MGLL | Monoacylglycerol lipase |
MAPT | Tau |
GSK3B | Glycogen synthase kinase-3β |
CDK5 | Cyclin-dependent-like kinase 5 |
MAOA | Amine oxidase A |
MAOB | Amine oxidase B |
PTGS2 | Cyclooxygenase-2 |
ApoE | Apolipoprotein E |
PPARG | Peroxisome proliferator-activated receptor-γ |
CREB | Camp response element binding |
HRH3 | Histamine 3 receptor |
MMP | Matrix metalloproteinase |
ABCA7 | Phospholipid-transporting ATPase ABCA7 |
ADAM10 | Disintegrin and metalloproteinase domain-containing protein 10 |
BIN1 | Bridging integrator 1 |
CB | Cannabinoid receptor |
FAAH | Fatty-acid amide hydrolase 1 |
Drug | Target | Kd | Prediction Result |
---|---|---|---|
1-[4-[(4-ethylpiperazin-1-yl)methyl]-3-(trifluoromethyl)phenyl]-3-[4-[6-(methylamino)pyrimidin-4-yl]oxyphenyl]urea | Cyclin-dependent-like kinase 5 | 1000 | √ |
DORAMAPIMOD | 2000 | √ | |
BMS-387032 | 740 | √ | |
2-(2-chlorophenyl)-5,7-dihydroxy-8-[(3S)-3-hydroxy-1-methyl-4-piperidyl]chromone | 110 | √ | |
JNJ-7706621 | 240 | √ | |
SELICICLIB | 1900 | √ | |
US8765747, 2 | 150 | √ | |
Carbacylamidophosphate, 1a | Acetylcholinesterase | 206,000 | √ |
Carbacylamidophosphate, 2a | 230,000 | √ | |
Carbacylamidophosphate, 3a | 1,100,000 | × | |
Carbacylamidophosphate, 4a | 554,000 | √ | |
Carbacylamidophosphate, 1b | 239,000 | √ | |
Carbacylamidophosphate, 2b | 3820 | × | |
Carbacylamidophosphate, 3b | 2,950,000 | × | |
Carbacylamidophosphate, 4b | 4,260,000 | √ |
Type | Drugs | Targets | Pos. Interactions | Neg. Interactions |
---|---|---|---|---|
Train Set | 8485 | 2861 | 15,648 | 15,648 |
Validation Set | 4376 | 2356 | 2236 | 9596 |
Test Set | 6708 | 2740 | 4471 | 19,192 |
Type | Datasets | Frequency | Scale of V |
---|---|---|---|
Drug (SMILES) | ChEMBL | 100 | 23,531 |
1500 | 2585 | ||
Target (amino acid sequence) | UniProt | 500 | 16,692 |
2000 | 4113 |
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2024 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
Share and Cite
Chen, Y.; Liang, X.; Du, W.; Liang, Y.; Wong, G.; Chen, L. Drug–Target Interaction Prediction Based on an Interactive Inference Network. Int. J. Mol. Sci. 2024, 25, 7753. https://doi.org/10.3390/ijms25147753
Chen Y, Liang X, Du W, Liang Y, Wong G, Chen L. Drug–Target Interaction Prediction Based on an Interactive Inference Network. International Journal of Molecular Sciences. 2024; 25(14):7753. https://doi.org/10.3390/ijms25147753
Chicago/Turabian StyleChen, Yuqi, Xiaomin Liang, Wei Du, Yanchun Liang, Garry Wong, and Liang Chen. 2024. "Drug–Target Interaction Prediction Based on an Interactive Inference Network" International Journal of Molecular Sciences 25, no. 14: 7753. https://doi.org/10.3390/ijms25147753