- 作者: Vijay Narayanasamy; Snehasis Mukhopadhyay; Mathew Palakal; David A. Potter
- 作者服務機構: School of Informatics, Department of Computer and Information Science, and Departments of Medicine and Biochemistry and Walther Oncology Center, Indiana University School of Medicine, Indiana University Purdue University Indianapolis, Ind., USA
- 中文摘要: --
- 英文摘要: Associations among biological objects such as genes,proteins, and drugs can be discovered automaticallyfrom the scientific literature.TransMiner is a system forfinding associations among objects by mining the Med-line database of the scientific literature.The direct asso-ciations among the objects are discovered based on theprinciple of co-occurrence in the form of an associationgraph.The principle of transitive closure is applied to theassociation graph to find potential transitive associa-tions.The potential transitive associations that are in-deed direct are discovered by iterative retrieval and min-ing of the Medline documents. Those associations thatare not found explicitly in the entire Medline databaseare transitive associations and are the candidates forhypothesis generation.The transitive associations wereranked based on the sum of weight of terms that co-occur with both the objects.The direct and transitiveassociations are visualized using a graph visualizationapplet. TransMiner was tested by finding associationsamong 56 breast cancer genes and among 24 objects inthe calpain signal transduction pathway. TransMinerwas also used to rediscover associations between mag-nesium and migraine.
- 中文關鍵字: --
- 英文關鍵字: --