PerspectiveFeatureDrug repurposing by integrated literature mining and drug-gene-disease triangulation

•Computational drug repositioning infers promising new candidates for existing drugs.•Existing approaches differ in the utilized data sources.•Combined analysis of heterogeneous databases on drugs, genes and diseases integrated with literature mining boosts the prediction power.•A new model for drug repositioning based on n-cluster editing is suggested.

Drug design is expensive, time-consuming and becoming increasingly complicated. Computational approaches for inferring potentially new purposes of existing drugs, referred to as drug repositioning, play an increasingly important part in current pharmaceutical studies. Here, we first summarize recent developments in computational drug repositioning and introduce the utilized data sources. Afterwards, we introduce a new data fusion model based on n-cluster editing as a novel multi-source triangulation strategy, which was further combined with semantic literature mining. Our evaluation suggests that utilizing drug-gene-disease triangulation coupled to sophisticated text analysis is a robust approach for identifying new drug candidates for repurposing.