Predicting drug-disease interactions by semi-supervised graph cut algorithm and three-layer data integration

Abstract

BackgroundPrediction of drug-disease interactions is promising for either drug repositioning or disease treatment fields. The discovery of novel drug-disease interactions, on one hand can help to find novel indictions for the approved drugs; on the other hand can provide new therapeutic approaches for the diseases. Recently, computational methods for finding drug-disease interactions have attracted lots of attention because of their far more higher efficiency and lower cost than the traditional wet experiment methods. However, they still face several challenges, such as the organization of the heterogeneous data, the performance of the model, and so on.MethodsIn this work, we present to hierarchically integrate the heterogeneous data into three layers. The drug-drug and disease-disease similarities are first calculated separately in each layer, and then the similarities from three layers are linearly fused into comprehensive drug similarities and disease similarities, which can then be used to measure the similarities between two drug-disease pairs. We construct a novel weighted drug-disease pair network, where a node is a drug-disease pair with known or unknown treatment relation, an edge represents the node-node relation which is weighted with the similarity score between two pairs. Now that similar drug-disease pairs are supposed to show similar treatment patterns, we can find the optimal graph cut of the network. The drug-disease pair with unknown relation can then be considered to have similar treatment relation with that within the same cut. Therefore, we develop a semi-supervised graph cut algorithm, SSGC, to find the optimal graph cut, based on which we can identify the potential drug-disease treatment interactions.ResultsBy comparing with three representative network-based methods, SSGC achieves the highest performances, in terms of both AUC score and the identification rates of true drug-disease pairs. The experiments with different integration strategies also demonstrate that considering several sources of data can improve the performances of the predictors. Further case studies on four diseases, the top-ranked drug-disease associations have been confirmed by KEGG, CTD database and the literature, illustrating the usefulness of SSGC.ConclusionsThe proposed comprehensive similarity scores from multi-views and multiple layers and the graph-cut based algorithm can greatly improve the prediction performances of drug-disease associations.

Identification of Drug-Disease Associations Using Information of Molecular Structures and Clinical Symptoms via Deep Convolutional Neural Network

“…On the other hand, disease burden is increasing globally due to the growth of population, outbreak of infectious disease and emergence of antibiotic resistance (Shameer et al, 2018). In order to circumvent this dilemma, drug repositioning has become a promising alternative strategy for drug research and development (Wu et al, 2017).…”

“…These computational approaches can be roughly divided into three mainstreams: drug-based, disease-based, and network-based. The former two are according to the assumption that drugs having similar structures/properties are inclined to be associated with diseases having similar pathogenesis/symptoms, and vice versa (Liu et al, 2016;Wu et al, 2017;Shameer et al, 2018). For example, Gottieb et al (2011) utilized multiple drug-drug and disease-disease similarity measures for the prediction of drug repurposing using the logistic regression classifier.…”

“…The last one is based on the principle of "guilt-byassociation" that drugs treating with same disease share structure/network properties and the diseases treated with the same drug also share phenotype/network properties (Wu et al, 2017). For instance, Zhao and Li (2012) defined a network-based gene closeness profile to relate drug to disease.…”

“…Yu et al (2016) represented a cluster method for prediction of new drug indications by using the identified drugs and disease modules based on the constructed drug network and disease network. Wu et al (2017) constructed a novel weighted drug-disease pair network, where a node is a drug-disease pair and a weighted edge represents the node-node relation. Then, a semi-supervised graph cut algorithm was adopted to identify the potential drug-disease treatment interactions.…”

SAEROF: an ensemble approach for large-scale drug-disease association prediction by incorporating rotation forest and sparse autoencoder deep neural network

“…The method incorporates information such as medicinal chemistry information and drug targets. To solve this problem, Wu et al proposed a semi-supervised graph cutting algorithm to find the optimal graph cutting to identify potential drug-disease associations, which is called SSGC 3 .…”