Pathway Curation Research Articles

WikiPathways: Pathway Editing for the People

The exponential growth of diverse types of biological data presents the research community with an unprecedented challenge and opportunity. The challenge is to stay afloat in the flood of biological data, keeping it as accessible, up-to-date, and integrated as possible. The opportunity is to cultivate new models of data curation and exchange that take advantage of direct participation by a greater portion of the community. This combination of challenge and opportunity is especially relevant to the task of collecting biological pathway information. Pathways are critical to understanding the functions of individual genes and proteins in terms of systems and processes that contribute to normal physiology and to disease. Each biological pathway must be hewn from a mass of biological information distributed across multiple publications and databases. The particular challenge of pathway curation is amplified, because pathways are often presented as static images that are not amenable to computation, integration, or data exchange. Furthermore, pathway experts are distributed throughout the world, and most have limited time to learn about complex databases that need their expertise. This challenge can be met by taking the opportunity to develop a new community-based model for pathway curation. One way to engage the community is with a wiki model, as exemplified by Wikipedia [1]. We see the potential for a wiki-based pathway curation resource, coupled with an embedded graphical pathway editing tool, to meet the growing challenge presented by the influx of biological data and to provide an innovative example of content curation by the biology community (Figure 1). Figure 1 Two Models for Managing Biological Data

Reactome – a knowledgebase of human biological pathways

Pathway curation is a powerful tool for systematically associating gene products with functions. Reactome (www.reactome.org) is a manually curated human pathway knowledgebase describing a wide range of biological processes in a computationally accessible manner. The core unit of the Reactome data model is the Reaction, whose instances form a network of biological interactions through entities that are consumed, produced, or act as catalysts. Entities are distinguished by their molecular identities and cellular locations. Set objects allow grouping of related entities. Curation is based on communication between expert authors and staff curators, facilitated by freely available data entry tools. Manually curated data are subjected to quality control and peer review by a second expert. Reactome data are released quarterly. At release time, electronic orthology inference performed on human data produces reaction predictions in 22 species ranging from mouse to bacteria. Cross-references to a large number of publicly available databases are attached, providing multiple entry points into the database. The Reactome Mart allows query submission and data retrieval from Reactome and across other databases. The SkyPainter tool provides visualization and statistical analysis of user supplied data, e.g. from microarray experiments. Reactome data are freely available in a number of data formats (e.g. BioPax, SBML).

Wnt pathway curation using automated natural language processing: combining statistical methods with partial and full parse for knowledge extraction

Wnt signaling is a very active area of research with highly relevant publications appearing at a rate of more than one per day. Building and maintaining databases describing signal transduction networks is a time-consuming and demanding task that requires careful literature analysis and extensive domain-specific knowledge. For instance, more than 50 factors involved in Wnt signal transduction have been identified as of late 2003. In this work we describe a natural language processing (NLP) system that is able to identify references to biological interaction networks in free text and automatically assembles a protein association and interaction map. A 'gold standard' set of names and assertions was derived by manual scanning of the Wnt genes website (http://www.stanford.edu/~rnusse/wntwindow.html) including 53 interactions involved in Wnt signaling. This system was used to analyze a corpus of peer-reviewed articles related to Wnt signaling including 3369 Pubmed and 1230 full text papers. Names for key Wnt-pathway associated proteins and biological entities are identified using a chi-squared analysis of noun phrases over-represented in the Wnt literature as compared to the general signal transduction literature. Interestingly, we identified several instances where generic terms were used on the website when more specific terms occur in the literature, and one typographic error on the Wnt canonical pathway. Using the named entity list and performing an exhaustive assertion extraction of the corpus, 34 of the 53 interactions in the 'gold standard' Wnt signaling set were successfully identified (64% recall). In addition, the automated extraction found several interactions involving key Wnt-related molecules which were missing or different from those in the canonical diagram, and these were confirmed by manual review of the text. These results suggest that a combination of NLP techniques for information extraction can form a useful first-pass tool for assisting human annotation and maintenance of signal pathway databases. The pipeline software components are freely available on request to the authors. dstates@umich.edu http://stateslab.bioinformatics.med.umich.edu/software.html.