KG-CNNDTI: a knowledge graph-enhanced prediction model for drug-target interactions and application in virtual screening of natural products against Alzheimer’s disease

Abstract

Accurate prediction of drug-target interactions (DTIs) plays a pivotal role in drug discovery, facilitating optimization of lead compounds, drug repurposing and elucidation of drug side effects. However, traditional DTI prediction methods are often limited by incomplete biological data and insufficient representation of protein features. In this study, we proposed KG-CNNDTI, a novel knowledge graph-enhanced framework for DTI prediction, which integrates heterogeneous biological information to improve model generalizability and predictive performance. The proposed model utilized protein embeddings derived from a biomedical knowledge graph via the Node2Vec algorithm, which were further enriched with contextualized sequence representations obtained from ProteinBERT. For compound representation, multiple molecular fingerprint schemes alongside the Uni-Mol pre-trained model were evaluated. The fused representations served as inputs to both classical machine learning models and a convolutional neural network-based predictor. Experimental evaluations across benchmark datasets demonstrated that KG-CNNDTI achieved superior performance compared to state-of-the-art methods, particularly in terms of Precision, Recall, F1-Score and area under the precision-recall curve (AUPR). Ablation analysis highlighted the substantial contribution of knowledge graph-derived features. Moreover, KG-CNNDTI was employed for virtual screening of natural products against Alzheimer's disease, resulting in 40 candidate compounds. 5 were supported by literature evidence, among which 3 were further validated in vitro assays.

Keywords

Drug-target interactions prediction / Knowledge graph / Drug screening / Alzheimer’s disease / Natural products

References

[1] New drug costs soar to $2. 6 billion. Nat Biotech. 2014; 32(12):1176.

[2] Zhang K, Yang X, Wang Y, et al.Artificial intelligence in drug development. Nat Med. 2025; 31(1):45-59. https://doi.org/10.1038/s41591-024-03434-4.

[3] Takebe T, Imai R, Ono S. The current status of drug discovery and development as originated in United States academia: the influence of industrial and academic collaboration on drug discovery and development. Clin Transl Sci. 2018; 11(6):597-606. https://doi.org/10.1111/cts.12577.

[4] Tang B, He F, Liu D, et al.AI-dided design of novel targeted covalent inhibitors against SARS-CoV-2. Biomolecules. 2022; 12(6):746. https://doi.org/10.3390/biom12060746.

[5] Ajmal A, Mahmood A, Hayat C, et al. Computer-assisted drug repurposing for thymidylate kinase drug target in monkeypox virus. Front Cell Infect Microbiol. 2023;13:1159389. https://doi.org/10.3389/fcimb.2023.1159389.

[6] Yu Z, Wu Z, Zhou M, et al. mtADENet: a novel interpretable method integrating multiple types of network-based inference approaches for prediction of adverse drug events. Comput Biol Med. 2024;168:107831. https://doi.org/10.1016/j.compbiomed.2023.107831.

[7] Xu L, Ru X, Song R. Application of machine learning for drug-target interaction prediction. Front Genet. 2021;12:680117. https://doi.org/10.3389/fgene.2021.680117.

[8] Dhakal A, McKay C, Tanner JJ, et al. Artificial intelligence in the prediction of protein-ligand interactions: recent advances and future directions. Brief Bioinform. 2022; 23(1):bbab476. https://doi.org/10.1093/bib/bbab476.

[9] Liao Q, Zhang Y, Chu Y, et al. Application of artificial intelligence in drug-target interactions prediction: a review. npj Biomed Innov. 2025; 2(1):1. https://doi.org/10.1038/s44385-024-00003-9.

[10] Xu Z, Guan L, Wang Y, et al. Discovery of a novel PLK1 inhibitor with high inhibitory potency using a combined virtual screening strategy. J Enzyme Inhib Med Chem. 2025; 40(1):2467798. https://doi.org/10.1080/14756366.2025.2467798.

[11] Sharma V, Wakode S, Kumar H. Chapter 2-structure-and ligand-based drug design:concepts, approaches, and challenges. Chemoinformatics and Bioinformatics in the Pharmaceutical Sciences. Amsterdam: Academic Press, 2021: 27-53.

[12] Ye Q, Hsieh CY, Yang Z, et al. A unified drug-target interaction prediction framework based on knowledge graph and recommendation system. Nat Commun. 2021; 12(1):6775. https://doi.org/10.1038/s41467-021-27137-3.

[13] Cheng F, Liu C, Jiang J, et al. Prediction of drug-target interactions and drug repositioning via network-based inference. PLoS Comput Biol. 2012; 8(5):e1002503. https://doi.org/10.1371/journal.pcbi.1002503.

[14] Wu Z, Cheng F, Li J, et al. SDTNBI: an integrated network and chemoinformatics tool for systematic prediction of drug-target interactions and drug repositioning. Brief Bioinform. 2017; 18(2):333-347. https://doi.org/10.1093/bib/bbw012.

[15] Wu Z, Lu W, Wu D, et al. In silico prediction of chemical mechanism of action via an improved network-based inference method. Brit J Pharmacol. 2016; 173(23):3372-3385. https://doi.org/10.1111/bph.13629.

[16] Wu Z, Ma H, Liu Z, et al. wSDTNBI: a novel network-based inference method for virtual screening. Chem Sci. 2022; 13(4): 1060-1079.

[17] Bi X, Zhang S, Ma W, et al. HiSIF-DTA: a hierarchical semantic information fusion framework for drug-target affinity prediction. IEEE J Biomed Health Inform. 2025; 29(3):1579-1590. https://doi.org/10.1109/JBHI.2023.3334239.

[18] Chu Z, Huang F, Fu H, et al. Hierarchical graph representation learning for the prediction of drug-target binding affinity. Inform Science. 2022; 613(C):507-523. https://doi.org/10.1016/j.ins.2022.09.043.

[19] Ma W, Zhang S, Li Z, et al. Predicting drug-target affinity by learning protein knowledge from biological networks. IEEE J Biomed Health Inform. 2023; 27(4):2128-2137. https://doi.org/10.1109/JBHI.2023.3240305.

[20] Shi H, Liu S, Chen J, et al. Predicting drug-target interactions using Lasso with random forest based on evolutionary information and chemical structure. Genomics. 2019; 111(6):1839-1852. https://doi.org/10.1016/j.ygeno.2018.12.007.

[21] Li F, Zhang Z, Guan J, et al. Effective drug-target interaction prediction with mutual interaction neural network. Bioinformatics. 2022; 38(14):3582-3589. https://doi.org/10.1093/bioinformatics/btac377.

[22] Peng Y, Zhao S, Zeng Z, et al. LGBMDF: a cascade forest framework with LightGBM for predicting drug-target interactions. Front Microbiol. 2022;13:1092467. https://doi.org/10.3389/fmicb.2022.1092467.

[23] Nguyen T, Le H, Quinn TP, et al. GraphDTA: predicting drug-target binding affinity with graph neural networks. Bioinformatics. 2021; 37(8):1140-1147. https://doi.org/10.1093/bioinformatics/btaa921.

[24] Chen L, Tan X, Wang D, et al. TransformerCPI: improving compound-protein interaction prediction by sequence-based deep learning with self-attention mechanism and label reversal experiments. Bioinformatics. 2020; 36(16):4406-4414. https://doi.org/10.1093/bioinformatics/btaa524.

[25] 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(6):e1007129. https://doi.org/10.1371/journal.pcbi.1007129.

[26] Sivley RM, Dou X, Meiler J, et al. Comprehensive analysis of constraint on the spatial distribution of missense variants in human protein structures. Am J Hum Genet. 2018; 102(3):415-426. https://doi.org/10.1016/j.ajhg.2018.01.017.

[27] Li N, Yang Z, Wang J, et al. Drug-target interaction prediction using knowledge graph embedding. iScience. 2024; 27(6):109393. https://doi.org/10.1016/j.isci.2024.109393.

[28] Grover A, Leskovec J.node2vec:scalable feature learning for networks. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 2016: 855-864. https://doi.org/10.1145/2939672.2939754.

[29] Brandes N, Ofer D, Peleg Y, et al. ProteinBERT: a universal deep-learning model of protein sequence and function. Bioinformatics. 2022; 38(8):2102-2110. https://doi.org/10.1093/bioinformatics/btac020.

[30] Zhou G, Gao Z, Ding Q, et al. Uni-mol:a universal 3d molecular representation learning framework. International Conference on Learning Representations 2023.

[31] Davis MI, Hunt JP, Herrgard S, et al. Comprehensive analysis of kinase inhibitor selectivity. Nat Biotechnol. 2011; 29(11):1046-1051. https://doi.org/10.1038/nbt.1990.

[32] Metz JT, Johnson EF, Soni NB, et al. Navigating the kinome. Nat Chem Biol. 2011; 7(4):200-202. https://doi.org/10.1038/nchembio.530.

[33] Ru J, Li P, Wang J, et al. TCMSP: a database of systems pharmacology for drug discovery from herbal medicines. J Cheminform. 2014; 6:13. https://doi.org/10.1186/1758-2946-6-13.

[34] Xu HY, Zhang YQ, Liu ZM, et al. ETCM: an encyclopaedia of traditional Chinese medicine. Nucleic Acids Res. 2019; 47(D1):D976-D982. https://doi.org/10.1093/nar/gky987.

[35] Huang L, Xie D, Yu Y, et al. TCMID 2. 0: a comprehensive resource for TCM. Nucleic Acids Res. 2018; 46(D1):D1117-D1120. https://doi.org/10.1093/nar/gkx1028.

[36] Kim S, Chen J, Cheng T, et al.PubChem 2025 update. Nucleic Acids Res.2024; 53(D1):D1516-D1525. https://doi.org/10.1093/nar/gkae1059.

[37] Gu Y, Yu Z, Wang Y, et al. admetSAR3. 0: a comprehensive platform for exploration, prediction and optimization of chemical ADMET properties. Nucleic Acids Res. 2024; 52(W1):W432-W438. https://doi.org/10.1093/nar/gkae298.

[38] Bickerton GR, Paolini GV, Besnard J, et al. Quantifying the chemical beauty of drugs. Nat Chem. 2012; 4(2):90-98. https://doi.org/10.1038/nchem.1243.

[39] Bian J, Zhang X, Zhang X, et al. MCANet: shared-weight-based MultiheadCrossAttention network for drug-target interaction prediction. Brief Bioinform. 2023; 24(2):bbad082. https://doi.org/10.1093/bib/bbad082.

[40] Huang K, Xiao C, Glass LM, et al. MolTrans: molecular Interaction transformer for drug-target interaction prediction. Bioinformatics. 2021; 37(6):830-836. https://doi.org/10.1093/bioinformatics/btaa880.

[41] Zhao Q, Zhao H, Zheng K, et al. HyperAttentionDTI: improving drug-protein interaction prediction by sequence-based deep learning with attention mechanism. Bioinformatics. 2022; 38(3):655-662. https://doi.org/10.1093/bioinformatics/btab715.

[42] Tao W, Lin X, Liu Y, et al. Bridging chemical structure and conceptual knowledge enables accurate prediction of compound-protein interaction. BMC Biol. 2024; 22(1):248. https://doi.org/10.1186/s12915-024-02049-y.

[43] Burley SK, Bhikadiya C, Bi C, et al. RCSB protein data bank: powerful new tools for exploring 3D structures of biological macromolecules for basic and applied research and education in fundamental biology, biomedicine, biotechnology, bioengineering and energy sciences. Nucleic Acids Res. 2020; 49(D1):D437-D451. https://doi.org/10.1093/nar/gkaa1038.

[44] Calsolaro V, Edison P. Neuroinflammation in Alzheimer’s disease: current evidence and future directions. Alzheimers Dement. 2016; 12(6):719-732. https://doi.org/10.1016/j.jalz.2016.02.010.

[45] Heneka MT, Kummer MP, Latz E. Innate immune activation in neurodegenerative disease. Nat Rev Immunol. 2014; 14(7):463-477. https://doi.org/10.1038/nri3705.

[46] Onyango IG, Jauregui GV, Čarná M, et al.Neuroinflammation in Alzheimer’s disease. Biomedicines. 2021; 9(5):524. https://doi.org/10.3390/biomedicines9050524.

[47] Rogers D, Hahn M. Extended-connectivity fingerprints. J Chem Inf Model. 2010; 50(5):742-754. https://doi.org/10.1021/ci100050t.

[48] Zheng Q, Wang X. Alzheimer’s disease: insights into pathology, molecular mechanisms, and therapy. Protein Cell. 2025; 16(2):83-120. https://doi.org/10.1093/procel/pwae026.

[49] Yue C, Chen B, Pan F, et al. TCnet: a novel strategy to predict target combination of Alzheimer’s disease via network-based methods. J Chem Inf Model. 2025; 65(7):3866-3878. https://doi.org/10.1021/acs.jcim.5c00172.

[50] Yang X, Xin Y, Gu Y, et al. Total alkaloids of Aconitum carmichaelii Debx alleviate cisplatin-induced acute renal injury by inhibiting inflammation and oxidative stress related to gut microbiota metabolism. Phytomedicine. 2024;135:156128. https://doi.org/10.1016/j.phymed.2024.156128.

[51] Nie H, Wang Z, Zhao W, et al. New nicotinic analogue ZY-1 enhances cognitive functions in a transgenic mice model of Alzheimer’s disease. Neurosci Lett. 2013; 537:29-34. https://doi.org/10.1016/j.neulet.2013.01.001.

[52] Wang L, Jiang Y, Yaseen A, et al. Steroidal alkaloids from the bulbs of Fritillaria pallidiflora Schrenk and their anti-inflammatory activity. Bioorg Chem. 2021;112:104845. https://doi.org/10.1016/j.bioorg.2021.104845.

[53] Liu T, Li Y, Sun J, et al. Gitogenin suppresses lung cancer progression by inducing apoptosis and autophagy initiation through the activation of AMPK signaling. Int Immunopharmacol. 2022;111:108806. https://doi.org/10.1016/j.intimp.2022.108806.

[54] Balkrishna A, Bhattacharya K, Shukla S, et al. Neuroprotection by polyherbal medicine divya-medha-vati against scopolamine-induced cognitive impairment through modulation of oxidative stress, acetylcholine activity, and cell signaling. Mol Neurobiol. 2024; 61(3):1363-1382. https://doi.org/10.1007/s12035-023-03601-7.

[55] Liu T, Wang S, Zhang Y, et al. TIWMFLP: two-tier interactive weighted matrix factorization and label propagation based on similarity matrix fusion for drug-disease association prediction. J Chem Inf Model. 2024; 64(22):8641-8654. https://doi.org/10.1021/acs.jcim.4c01589.

[56] Wang S, Liu T, Ren C, et al. Predicting potential small molecule-miRNA associations utilizing truncated schatten p-norm. Brief Bioinform. 2023; 24(4):bbad234. https://doi.org/10.1093/bib/bbad234.

[57] Zhang L, Wang CC, Zhang Y, et al. GPCNDTA: prediction of drug-target binding affinity through cross-attention networks augmented with graph features and pharmacophores. Comput Biol Med. 2023;166:107512. https://doi.org/10.1016/j.compbiomed.2023.107512.

[58] Li J, Bi X, Ma W, et al. MHAN-DTA: a multiscale hybrid attention network for drug-target affinity prediction. IEEE J Biomed Health Inform 2024. 2024; 2024:1-21. https://doi.org/10.1109/JBHI.2024.3518619.

[59] Liu Z, Chen Q, Lan W, et al. SSLDTI: a novel method for drug-target interaction prediction based on self-supervised learning. Artif Intell Med. 2024;149:102778. https://doi.org/10.1016/j.artmed.2024.102778.

[60] Peng W, Chen T, Liu H, et al. Improving drug response prediction based on two-space graph convolution. Comput Biol Med. 2023;158:106859. https://doi.org/10.1016/j.compbiomed.2023.106859.

[61] Lan W, Zhou G, Chen Q, et al. Contrastive clustering learning for multi-behavior recommendation. ACM Trans Inf Syst. 2024; 43(1): Article 18. https://doi.org/10.1145/3698192.

[62] Lan W, Tang Z, Liu M, et al. The large language models on biomedical data analysis: a survey. IEEE J Biomed Health Inform. 2025; 29(6):4486-4497. https://doi.org/10.1109/JBHI.2025.3530794.