STNN-DDI: A Substructure-aware Tensor Neural Network to Predict Drug-Drug Interactions

Computational prediction of multiple-type drug-drug interaction (DDI) helps reduce unexpected side effects in poly-drug treatments. Although existing computational approaches achieve inspiring results, they ignore that the action of a drug is mainly caused by its chemical substructures. In addition, their interpretability is still weak. In this paper, by supposing that the interactions between two given drugs are caused by their local chemical structures (substructures) and their DDI types are determined by the linkages between different substructure sets, we design a novel Substructure-ware Tensor Neural Network model for DDI prediction (STNN-DDI). The proposed model learns a 3-D tensor of (substructure, interaction type, substructure) triplets, which characterizes a substructure-substructure interaction (SSI) space. According to a list of predefined substructures with specific chemical meanings, the mapping of drugs into this SSI space enables STNN-DDI to perform the multiple-type DDI prediction in both transductive and inductive scenarios in a unified form with an explicable manner. The comparison with deep learning-based state-of-the-art baselines demonstrates the superiority of STNN-DDI with the significant improvement of AUC, AUPR, Accuracy, and Precision. More importantly, case studies illustrate its interpretability by both revealing an important substructure pair across drugs regarding a DDI type of interest and uncovering interaction type-specific substructure pairs in a given DDI. In summary, STNN-DDI provides an effective approach to predicting DDIs as well as explaining the interaction mechanisms among drugs. Source code is freely available at https://github.com/zsy9/STNN-DDI.

[1]  Tamara G. Kolda,et al.  All-at-once Optimization for Coupled Matrix and Tensor Factorizations , 2011, ArXiv.

[2]  Hisashi Kashima,et al.  Tensor factorization using auxiliary information , 2011, Data Mining and Knowledge Discovery.

[3]  Stephan Günnemann,et al.  Introduction to Tensor Decompositions and their Applications in Machine Learning , 2017, ArXiv.

[4]  Ping Zhang,et al.  Multitask Dyadic Prediction and Its Application in Prediction of Adverse Drug-Drug Interaction , 2017, AAAI.

[5]  Reza Safdari,et al.  Computational prediction of drug-drug interactions based on drugs functional similarities , 2017, J. Biomed. Informatics.

[6]  Jimeng Sun,et al.  Rubik: Knowledge Guided Tensor Factorization and Completion for Health Data Analytics , 2015, KDD.

[7]  Yongguo Liu,et al.  MTMA: Multi-task multi-attribute learning for the prediction of adverse drug-drug interaction , 2020, Knowl. Based Syst..

[8]  Yanli Wang,et al.  Predicting drug–drug interactions through drug structural similarities and interaction networks incorporating pharmacokinetics and pharmacodynamics knowledge , 2017, Journal of Cheminformatics.

[9]  Nikos D. Sidiropoulos,et al.  Tensor Decomposition for Signal Processing and Machine Learning , 2016, IEEE Transactions on Signal Processing.

[10]  Tamara G. Kolda,et al.  Tensor Decompositions and Applications , 2009, SIAM Rev..

[11]  Jae Yong Ryu,et al.  Deep learning improves prediction of drug–drug and drug–food interactions , 2018, Proceedings of the National Academy of Sciences.

[12]  N. Shah,et al.  Performance of Pharmacovigilance Signal‐Detection Algorithms for the FDA Adverse Event Reporting System , 2013, Clinical pharmacology and therapeutics.

[13]  Sajid Ahmed,et al.  FRnet-DTI: Deep convolutional neural network for drug-target interaction prediction , 2020, Heliyon.

[14]  Yanlin Chen,et al.  SFLLN: A sparse feature learning ensemble method with linear neighborhood regularization for predicting drug-drug interactions , 2019, Inf. Sci..

[15]  L B Sheiner,et al.  Simultaneous Modeling of Pharmacokinetics and Pharmacodynamics with a Nonparametric Pharmacodynamic Model , 1984, Clinical pharmacology and therapeutics.

[16]  Meng Wang,et al.  Predicting Rich Drug-Drug Interactions via Biomedical Knowledge Graphs and Text Jointly Embedding , 2017, ArXiv.

[17]  Wen Zhang,et al.  A multimodal deep learning framework for predicting drug-drug interaction events , 2020, Bioinform..

[18]  Jason Lazarou,et al.  Incidence of Adverse Drug Reactions in Hospitalized Patients , 1999 .

[19]  D. Bojanic,et al.  Keynote review: in vitro safety pharmacology profiling: an essential tool for successful drug development. , 2005, Drug discovery today.

[20]  R. Altman,et al.  Data-Driven Prediction of Drug Effects and Interactions , 2012, Science Translational Medicine.

[21]  Ali Masoudi-Nejad,et al.  Synthetic repurposing of drugs against hypertension: a datamining method based on association rules and a novel discrete algorithm , 2020, BMC Bioinformatics.

[22]  Chihyun Park,et al.  Novel deep learning model for more accurate prediction of drug-drug interaction effects , 2019, BMC Bioinformatics.

[23]  Feng Liu,et al.  Predicting potential drug-drug interactions by integrating chemical, biological, phenotypic and network data , 2017, BMC Bioinformatics.

[24]  Jiangtao Wang,et al.  MLRDA: A Multi-Task Semi-Supervised Learning Framework for Drug-Drug Interaction Prediction , 2019, IJCAI.

[25]  Hans-Peter Kriegel,et al.  A Three-Way Model for Collective Learning on Multi-Relational Data , 2011, ICML.

[26]  Siu-Ming Yiu,et al.  Predicting and understanding comprehensive drug-drug interactions via semi-nonnegative matrix factorization , 2018, BMC Systems Biology.

[27]  Jimeng Sun,et al.  CASTER: Predicting Drug Interactions with Chemical Substructure Representation , 2019, AAAI.

[28]  Jianying Hu,et al.  Label Propagation Prediction of Drug-Drug Interactions Based on Clinical Side Effects , 2015, Scientific Reports.

[29]  R. Krauss,et al.  When good drugs go bad , 2007, Nature.

[30]  Changiz Eslahchi,et al.  ISCMF: Integrated similarity-constrained matrix factorization for drug–drug interaction prediction , 2020, Network Modeling Analysis in Health Informatics and Bioinformatics.

[31]  Diane L. Seger,et al.  Viewpoint Paper: Tiering Drug-Drug Interaction Alerts by Severity Increases Compliance Rates , 2009, J. Am. Medical Informatics Assoc..

[32]  R. Sharan,et al.  INDI: a computational framework for inferring drug interactions and their associated recommendations , 2012, Molecular systems biology.

[33]  H. Sobhanam,et al.  Addressing cold start problem in recommender systems using association rules and clustering technique , 2013, 2013 International Conference on Computer Communication and Informatics.

[34]  F. L. Hitchcock.  The Expression of a Tensor or a Polyadic as a Sum of Products , 1927 .