Network-based inference methods for drug repositioning

Abstract

Mining potential drug-disease associations can speed up drug repositioning for pharmaceutical companies. Previous computational strategies focused on prior biological information for association inference. However, such information may not be comprehensively available and may contain errors. Different from previous research, two inference methods, ProbS and HeatS, were introduced in this paper to predict direct drug-disease associations based only on the basic network topology measure. Bipartite network topology was used to prioritize the potentially indicated diseases for a drug. Experimental results showed that both methods can receive reliable prediction performance and achieve AUC values of 0.9192 and 0.9079, respectively. Case studies on real drugs indicated that some of the strongly predicted associations were confirmed by results in the Comparative Toxicogenomics Database (CTD). Finally, a comprehensive prediction of drug-disease associations enables us to suggest many new drug indications for further studies.

Figure 1

The flowchart of the two methods ProbS (a) and HeatS (b). The cylinder objects and the ellipse objects mean drugs and diseases, respectively. This figure is inspired by [33].

Figure 2

Drug-disease association network. Red rectangles and yellow rectangles indicate drugs and diseases, respectively. The bipartite network is generated by using 1933 experimentally verified associations between drugs and diseases. This network is prepared by Cytoscape (http://www.cytoscape.org/).

Figure 3

Degree distributions of drugs and diseases in the drug-disease network. (a) shows the histograms of the degree distributions of drugs. (b) shows the histograms of the degree distributions of diseases.