Automatic pathway information extraction from biomedical literature

Automatically extracting protein pathway information from the vast amount of published literature helps us understand diseases and discover new drug targets. In this paper, we propose a syntactic rule-based method for extracting pathway relation information from Medline abstracts. This method takes into account comprehensive dependency types and negative connectives when extracting relations from syntactic analysis results. The benchmarking results demonstrate the effectiveness of our method. © 2013 by Binary Information Press.