Ontology-based information extraction of regulatory networks from scientific articles with case studies for Escherichia coli

The amount of scientific papers in the Molecular Biology field has experienced an enormous growth in the last years, prompting the need of developing automatic Information Extraction (IE) systems. This work is a first step towards the ontology-based domain-independent generalization of a system that identifies Escherichia coli regulatory networks. First, a domain ontology based on the RegulonDB database was designed and populated. After that, the steps of the existing IE system were generalized to use the knowledge contained in the ontology, so that it could be potentially applied to other domains. The resulting system has been tested both with abstract and full articles that describe regulatory interactions for E. coli, obtaining satisfactory results.

► Ontologies may be used to support Information Extraction systems.
► A first step towards the generalization of an IE system focused on the detection of regulatory networks has been made.
► The domain ontology that guides the IE follows the structure of RegulonDB and has been populated with its content.