To efficiently save cost and reduce risk in drug research and development, there is a pressing demand to develop in silico methods to predict drug sensitivity to cancer cells. With the exponentially increasing number of multi-omics data derived from high-throughput techniques, machine learning-based methods have been applied to the prediction of drug sensitivities. However, these methods have drawbacks either in the interpretability of the mechanism of drug action or limited performance in modeling drug sensitivity. In this paper, we presented a pathway-guided deep neural network (DNN) model to predict the drug sensitivity in cancer cells. Biological pathways describe a group of molecules in a cell that collaborates to control various biological functions like cell proliferation and death, thereby abnormal function of pathways can result in disease. To take advantage of the excellent predictive ability of DNN and the biological knowledge of pathways, we reshaped the canonical DNN structure by incorporating a layer of pathway nodes and their connections to input gene nodes, which makes the DNN model more interpretable and predictive compared to canonical DNN. We have conducted extensive performance evaluations on multiple independent drug sensitivity data sets and demonstrated that our model significantly outperformed the canonical DNN model and eight other classical regression models. Most importantly, we observed a remarkable activity decrease in disease-related pathway nodes during forward propagation upon inputs of drug targets, which implicitly corresponds to the inhibition effect of disease-related pathways induced by drug treatment on cancer cells. Our empirical experiments showed that our method achieves pharmacological interpretability and predictive ability in modeling drug sensitivity in cancer cells. The web server, the processed data sets, and source codes for reproducing our work are available at .

Drug research and development is a time-consuming and high-cost task, pressing an urgent demand to identify novel indications of approved drugs, referred to as drug repositioning, which provides an economical and efficient way for drug discovery. With increasing volumes of large-scale chemical, genomic, and pharmacological data sets generated by the high-throughput technique, it is crucial to develop systematic and rational computational approaches to identify new indications of approved drugs. In this paper, we introduce HNet-DNN, which utilizes a deep neural network (DNN), to predict new drug–disease associations based on the features extracted from the drug–disease heterogeneous network. Instead of the straightforward concatenation of chemical and phenotypic features as the input of DNN, we used these raw features of drugs and diseases to construct a drug–drug similarity network and a disease–disease similarity network, and then built a drug–disease heterogeneous network by integrating known drug–disease associations. Subsequently, we extracted topological features for drug–disease associations from the heterogeneous network and used them to train a DNN model. Our intensive performance evaluations demonstrated that HNet-DNN effectively exploits the features of the heterogeneous network to boost the predictive performance of drug–disease associations. Compared with a couple of typical classifiers and competitive approaches, our method not only achieved state-of-the-art performance but also effectively alleviated the overfitting problem. Moreover, we ran HNet-DNN to predict new drug–disease associations and carried out case studies to verify the effectiveness of our method.

No abstract