Evidence based disease network construction towards drug repositioning

Drug repositioning is one of emerging approaches dedicated to find alternative usages of existing drugs efficiently and economically, especially with the advance in computational technology. The current progress made for computational drug repositioning is primarily focusing on informatics approach development/improvement or exploration on different type of data in order to identify possible drug candidates. Comparing to the existing studies, we proposed a novel method for constructing the disease based network by applying data extracted from the Semantic MEDLINE. Phenotypical associations (disease-disease associations) can be identified from this network, which can drive drug repositioning study by targeting on specific domain. In this paper, we successfully demonstrated the capability of the disease based network in hidden phenotypical association discovery to support drug repositioning in case studies.