Literature mining, gene-set enrichment and pathway analysis for target identification in Behçet's disease.

Objective. To use literature mining to catalogue Behcet's associated genes, and advanced computational methods to improve the understanding of the pathways and signalling mechanisms that lead to the typical clinical characteristics of Behcet's patients. To extend this technique to identify potential treatment targets for further experimental validation. Methods. Text mining methods combined with gene enrichment tools, pathway analysis and causal analysis algorithms. Results. This approach identified 247 human genes associated with Behcet's disease and the resulting disease map, comprising 644 nodes and 19220 edges, captured important details of the relationships between these genes and their associated pathways, as described in diverse data repositories. Pathway analysis has identified how Behcet's associated genes are likely to participate in innate and adaptive immune responses. Causal analysis algorithms have identified a number of potential therapeutic strategies for further investigation. Conclusion. Computational methods have captured pertinent features of the prominent disease characteristics presented in Behcet's disease and have highlighted NOD2, ICOS and IL18 signalling as potential therapeutic strategies.