A novel paradigm for potential drug-targets discovery: quantifying relationships of enzymes and cascade interactions of neighboring biological processes to identify drug-targets.

Target discovery is the most crucial step in a modern drug discovery development. Our objective in this study is to propose a novel paradigm for a better discrimination of drug-targets and non-drug-targets with minimum disruptive side-effects under a biological pathway context. We introduce a novel metric, namely, "pathway closeness centrality", for each gene that jointly considers the relationships of its neighboring enzymes and cross-talks of biological processes, to evaluate its probability of being a drug-target. This metric could distinguish drug-targets with non-drug-targets. Genes with lower pathway closeness centrality values are prone to play marginal roles in biological processes and have less lethality risk, but appear to have tissue-specific expressions. Compared with traditional metrics, our method outperforms degree, betweenness and bridging centrality under the human pathway context. Analysis of the existing top 20 drugs with the most disruptive side-effects indicates that pathway closeness centrality is an appropriate index to predict the probability of the occurrence of adverse pharmacological effects. Case studies in prostate cancer and type 2 diabetes mellitus indicate that the pathway closeness centrality metric could distinguish likely drug-targets well from human pathways. Thus, our method is a promising tool to aid target identification in drug discovery.