Seekbio: Retrieval of Spatial Relations for System Biology

Introduction

For decades, it was thought that more an organism was complex, more the number of genes they contain had to be high. At the completion of the Human Genome Project, it was found that living systems were having approximately the same number of genes coding for proteins (about 20,000) and these assessments resulted in biology’s big bang undergoing a paradigmatic change to address biological complexity. Accordingly, the universe of biologists was changing and the challenge of unravelling complexity of living organisms was stipulated on the spatio-temporal variety of the biological components and their relations. Countless examples show that identical molecules could play different roles depending of the relations they may have with different partners embedded in different processes, at different time and/or localization; let mention, muscle differentiation that is orchestrated by the differential localization of a molecular species (Misca et al., 2001) or molecular gradients that regulate the developing eye polarity (Strutt, 1999). To address such intricate networks that account for the complexity of living systems, systems biology is an emerging field that aims at understanding such dynamic interactions from the knowledge of their components and the relations between these components. Among main issues in systems biology, knowledge on entities spatial relations is of importance to assess the topology of biological networks; in this perspective, mining data and texts could afford specific clues. To address this issue, we examine the use of contextual exploration method (Jouis, 1993; Jouis, 2007) to develop extraction rules that can retrieve information on relations between biological entities in scientific literature. To achieve this, we propose SEEKbio tool. Among further uses, SEEKbio could be plugged at Pubmed (Srinivasan, P., 2001) output as an interface between results of PubMed query and articles selection following spatial relationships requests. These spatial relationships requests would consist of terms graph of the original query and location relations put by the user in the SEEKbio graphs editor. The built graph would act as a filter to select relevant articles from a spatial point of view (see Figure 1). Enabling this strategy should return selected documents in which relations arguments could be added automatically.

Figure 1.

Utilisation of SEEKbio for collecting all relevant information through Pubmed applied to Medline. As a first step, the user makes a query in the form of a Boolean expression of terms. In step (2), the request is submitted to Pubmed via Medline database. Quite often, the user gets out of a hundred references. This result is not exploitable. So, in a third step, the user specifies his request in the form of a graph in the system SEEKbio. This graph represents a hierarchy of spatial terms of the original request. The graph serves as a filter. With the contextual exploration rules, SEEKbio detects texts in which appear the spatial relationships expressed by the graph. Finally, in Step 4, we get a few abstracts which verify the conditions expressed by the spatial hierarchy.

Related Works

Two different approaches to extracting relationships from biological texts are used (Ananiadou, S., Kell, D. B. & J. Tsujii., 2006).