Department of Health Information Technology, School of Management and Medical Informatics, Tabriz University of Medical Sciences, Tabriz, 51656/65811, Iran.
Department of Pharmaceutical Sciences, College of Pharmacy, Nova Southeastern University, Fort Lauderdale, FL, 33328, USA.
Sci Rep. 2023 Apr 15;13(1):6184. doi: 10.1038/s41598-023-33271-3.
Drug combinations can be the prime strategy for increasing the initial treatment options in cancer therapy. However, identifying the combinations through experimental approaches is very laborious and costly. Notably, in vitro and/or in vivo examination of all the possible combinations might not be plausible. This study presented a novel computational approach to predicting synergistic drug combinations. Specifically, the deep neural network-based binary classification was utilized to develop the model. Various physicochemical, genomic, protein-protein interaction and protein-metabolite interaction information were used to predict the synergy effects of the combinations of different drugs. The performance of the constructed model was compared with shallow neural network (SNN), k-nearest neighbors (KNN), random forest (RF), support vector machines (SVMs), and gradient boosting classifiers (GBC). Based on our findings, the proposed deep neural network model was found to be capable of predicting synergistic drug combinations with high accuracy. The prediction accuracy and AUC metrics for this model were 92.21% and 97.32% in tenfold cross-validation. According to the results, the integration of different types of physicochemical and genomics features leads to more accurate prediction of synergy in cancer drugs.
药物联合治疗可以成为增加癌症治疗初始治疗选择的主要策略。然而，通过实验方法来确定这些联合治疗方案非常费力且昂贵。值得注意的是，体外和/或体内检查所有可能的组合可能并不合理。本研究提出了一种预测协同药物组合的新计算方法。具体来说，利用基于深度神经网络的二进制分类来开发模型。利用各种物理化学、基因组、蛋白质-蛋白质相互作用和蛋白质-代谢物相互作用信息来预测不同药物组合的协同效应。所构建模型的性能与浅层神经网络（SNN）、k-最近邻（KNN）、随机森林（RF）、支持向量机（SVMs）和梯度提升分类器（GBC）进行了比较。基于我们的研究结果，发现所提出的深度神经网络模型能够以高精度预测协同药物组合。在十折交叉验证中，该模型的预测准确性和 AUC 指标分别为 92.21%和 97.32%。结果表明，整合不同类型的物理化学和基因组特征可更准确地预测癌症药物的协同作用。
