Computational Approach to Biological Validation of Protein-Protein Interactions Discovered using Literature Mining

In this paper we describe a computational technique to validate protein-protein interactions discovered using BioMap, a literature mining system. Two experiments were carried out to validate the interactions pairs. In this study, we used a total of 32,693 documents with 193,738 sentences related to multiple sclerosis from the Medline literature database. From these sentences, BioMap found 1004 explicit and 15,000 implicit protein interaction pairs of which 271 explicit and 16 implicit interactions were validated using the proposed techniques. These results indicate that the proposed methods can be effectively used for biological verification of protein-protein interactions that are obtained through literature mining.