Pathway construction and extension using natural language processing

Abstract

Construction and maintenance of signaling pathway is a time-consuming and labor-intensive task. In addition, integration of various pathways is also ineffective since several markup languages are used to express pathways. To overcome these limitation, automatic pathway construction and extension with a standard format may provide a solution. The proposed approach has constructed a gold standard corpus that describes the signaling pathways, and it has been used to training and evaluating the automatic pathway construction and extension. Moreover, a standard format to express the signaling pathways has been developed and has been used to express the previous major 10 signaling pathways. An effective visualization tool has been also developed for the standardized format as well. The visualization tool can help to construct pathways and extend the current pathways using all articles in PubMed. © 2013 Springer-Verlag Berlin Heidelberg.