藥物副作用或藥物不良反應一直是主要的藥學問題之一，並且是使用藥物治療失敗和藥物戒斷的主要原因。 此外，當多種藥物並服時，更可能誘發藥物相互作用，這可能導致更多的副作用風險。 在本文中，我們開發了一種新方法，使用深度圖卷積學習和轉移學習，基於化學結構和ADME（吸收，分佈，代謝和排泄）信息預測多種候選藥物的藥物副作用。

Drug side-effects or adverse drug reactions, always are one of major public health concerns that remains to be main causes of drug failure and drug withdrawal in medical treatment. In addition, nowadays, when it comes to taking multiple drugs, it is more likely to induce drug interactions that may lead to more risks in side effects [1]. In this paper we develop a new method to predict drug-drug side effects of multiple drug candidate molecules based on their chemical structures and their ADME(Absorption, Distribution, Metabolism, and Excretion) information using deep graph convolution learning and transfer learning.

摘要...............................................i
Abstract........................................ii
List of Tables................................vi
List of Figures..............................vii
Introduction...................................1
1.1Pharmacodynamic and pharmacokinetic lead to drug effect................1
1.2Drug side effects are related to its chemical structure.........................4
1.2.1The relationship between the physical properties of drugs and theefficacy of drugs................5
1.2.2The effect of drug dissociation on drug efficacy................7
1.2.3The effect of the functional group of the drug structure on the phys-ical properties and efficacy of the drug................8
1.2.4Effects of electron cloud density and three-dimensional structureon drug efficacy................10
1.2.5The relationship between the three-dimensional structure of drugsand drug efficacy...............10
1.2.6Conformation and biological activity of drugs................11
1.2.7Graph Convolutional Network Recap..............................16
Motivation.................................................18
Related Work............................................20
2.1Chemical-Based Methods...................20
2.2Biological-Based and Phenotypic-Based Methods................21
2.3Integrated-Information-Based Methods.................................21
2.4Drug combination modeling................22
2.5Neural networks on graphs.................22
Methodology.................................24
3.1Datasets...................................24
3.1.1DrugBank...............................24
3.1.2Drug-drug interaction and side effect data................................25
3.1.3Molecular structure................26
3.1.4Drug property.........................26
3.2Methods....................................28
3.2.1Graph convolutional with transfer learning approach................28
3.2.2Chemical encoder..............................29
3.2.3Network decoder...............................30
3.3Training of the Models..........................31
Experiments and Results...........................33
4.1The side effect prediction.....................34
4.1.1General prediction..............................34
4.1.2Comparison with other methods........35
4.2Case study............................................37
Conclusion and Future Work.....................39
5.1Conclusion and Future Work.................39
References.................................................41

1.Shah, B.M. and E.R. Hajjar, Polypharmacy, adverse drug reactions, and geriatric syndromes. Clin Geriatr Med, 2012. 28(2): p. 173-86.
2.Yan, J., et al., Network approaches to systems biology analysis of complex disease: integrative methods for multi-omics data. Brief Bioinform, 2018. 19(6): p. 1370-1381.
3.Di, L. and E.H. Kerns, Application of pharmaceutical profiling assays for optimization of drug-like properties. Curr Opin Drug Discov Devel, 2005. 8(4): p. 495-504.
4.Di, L. and E.H. Kerns, Chapter 1 - Introduction, in Drug-Like Properties (Second Edition), L. Di and E.H. Kerns, Editors. 2016, Academic Press: Boston. p. 1-3.
5.Mao, F., et al., Chemical Structure-Related Drug-Like Criteria of Global Approved Drugs. Molecules, 2016. 21(1): p. 75.
6.Welling, P.G., Interactions affecting drug absorption. Clin Pharmacokinet, 1984. 9(5): p. 404-34.
7.Song, I., et al., Pharmacokinetics of dolutegravir when administered with mineral supplements in healthy adult subjects. J Clin Pharmacol, 2015. 55(5): p. 490-6.
8.Krejsa, C.M., et al., Predicting ADME properties and side effects: the BioPrint approach. Curr Opin Drug Discov Devel, 2003. 6(4): p. 470-80.
9.Krizhevsky, A., I. Sutskever, and G.E. Hinton, ImageNet classification with deep convolutional neural networks, in Proceedings of the 25th International Conference on Neural Information Processing Systems - Volume 1. 2012, Curran Associates Inc.: Lake Tahoe, Nevada. p. 1097-1105.
10.Kipf, T.N. and M. Welling Semi-Supervised Classification with Graph Convolutional Networks. arXiv e-prints, 2016.
11.van den Berg, R., T.N. Kipf, and M. Welling Graph Convolutional Matrix Completion. arXiv e-prints, 2017.
12.Pauwels, E., V. Stoven, and Y. Yamanishi, Predicting drug side-effect profiles: a chemical fragment-based approach. BMC Bioinformatics, 2011. 12: p. 169.
13.Chen, L., et al., Predicting chemical toxicity effects based on chemical-chemical interactions. PLoS One, 2013. 8(2): p. e56517.
14.Huang, L.C., X. Wu, and J.Y. Chen, Predicting adverse side effects of drugs. BMC Genomics, 2011. 12 Suppl 5: p. S11.
15.Campillos, M., et al., Drug target identification using side-effect similarity. Science, 2008. 321(5886): p. 263-6.
16.Yamanishi, Y., E. Pauwels, and M. Kotera, Drug side-effect prediction based on the integration of chemical and biological spaces. J Chem Inf Model, 2012. 52(12): p. 3284-92.
17.Li, J., et al., A survey of current trends in computational drug repositioning. Brief Bioinform, 2016. 17(1): p. 2-12.
18.Han, K., et al., Synergistic drug combinations for cancer identified in a CRISPR screen for pairwise genetic interactions. Nat Biotechnol, 2017. 35(5): p. 463-474.
19.Lewis, R., et al., Synergy Maps: exploring compound combinations using network-based visualization. J Cheminform, 2015. 7: p. 36.
20.Shi, J.Y., et al., Predicting combinative drug pairs towards realistic screening via integrating heterogeneous features. BMC Bioinformatics, 2017. 18(Suppl 12): p. 409.
21.Gilmer, J., et al. Neural Message Passing for Quantum Chemistry. arXiv e-prints, 2017.
22.Schlichtkrull, M., et al. Modeling Relational Data with Graph Convolutional Networks. arXiv e-prints, 2017.
23.Ying, R., et al. Graph Convolutional Neural Networks for Web-Scale Recommender Systems. arXiv e-prints, 2018.
24.Wu, Z., et al. A Comprehensive Survey on Graph Neural Networks. arXiv e-prints, 2019.
25.Li, P., Y. Fu, and Y. Wang, Network based approach to drug discovery: a mini review. Mini Rev Med Chem, 2015. 15(8): p. 687-95.
26.Boezio, B., et al., Network-based Approaches in Pharmacology. Mol Inform, 2017. 36(10).
27.Wishart, D.S., et al., DrugBank 5.0: a major update to the DrugBank database for 2018. Nucleic Acids Res, 2018. 46(D1): p. D1074-D1082.
28.Zitnik, M., M. Agrawal, and J. Leskovec, Modeling polypharmacy side effects with graph convolutional networks. Bioinformatics, 2018. 34(13): p. i457-i466.
29.Kim, S., et al., PubChem 2019 update: improved access to chemical data. Nucleic Acids Res, 2019. 47(D1): p. D1102-D1109.
30.Dong, J., et al., ADMETlab: a platform for systematic ADMET evaluation based on a comprehensively collected ADMET database. J Cheminform, 2018. 10(1): p. 29.
31.Kuhn, M., et al., The SIDER database of drugs and side effects. Nucleic Acids Res, 2016. 44(D1): p. D1075-9.
32.Vinyals, O., S. Bengio, and M. Kudlur Order Matters: Sequence to sequence for sets. arXiv e-prints, 2015.
33.Nickel, M., V. Tresp, and H.-P. Kriegel, Factorizing YAGO: scalable machine learning for linked data, in Proceedings of the 21st international conference on World Wide Web. 2012, ACM: Lyon, France. p. 271-280.
34.Perozzi, B., R. Al-Rfou, and S. Skiena DeepWalk: Online Learning of Social Representations. arXiv e-prints, 2014.
35.Jaekoo Lee, H.K. and S.Y. Jongsun Lee Transfer Learning for Deep Learning on Graph-Structured Data. 2017.
36.Eraslan, G., et al., Deep learning: new computational modelling techniques for genomics. Nat Rev Genet, 2019. 20(7): p. 389-403.
37.Walder, A. and P. Baumann, Mood stabilizer therapy and pravastatin: higher risk for adverse skin reactions? Acta Medica (Hradec Kralove), 2009. 52(1): p. 15-8.
38.Cadeddu, G., et al., Clozapine toxicity due to a multiple drug interaction: a case report. J Med Case Rep, 2015. 9: p. 77.
39.Friedrich, M.E., et al., Drug-Induced Liver Injury during Antidepressant Treatment: Results of AMSP, a Drug Surveillance Program. Int J Neuropsychopharmacol, 2016. 19(4).
40.Cronin, S. and P.H. Chandrasekar, Safety of triazole antifungal drugs in patients with cancer. J Antimicrob Chemother, 2010. 65(3): p. 410-6.
41.Kotlinska-Lemieszek, A., P. Klepstad, and D.F. Haugen, Clinically significant drug-drug interactions involving opioid analgesics used for pain treatment in patients with cancer: a systematic review. Drug Des Devel Ther, 2015. 9: p. 5255-67.
42. https://www.sciencedirect.com/topics/medicine-and-dentistry/adme