BMC Bioinformatics

official impact factor 3.03

Open Access Software

GenoLink: a graph-based querying and browsing system for investigating the function of genes and proteins

Patrick Durand1,5*, Laurent Labarre1,2, Alain Meil1, Jean-Louis Divo11, Yves Vandenbrouck3,6, Alain Viari4 and Jérôme Wojcik1,7

Author Affiliations

1 Hybrigenics SA, 3–5 Impasse Reille, 75014 Paris, France

2 AGC, UMR CNRS 8030 – Genoscope, 2 rue Gaston Crémieux, 91000 Evry, France

3 Genome Express, 11 Chemin des Prés, 38944 Meylan, France

4 INRIA Rhône-Alpes, 655 Avenue de l'Europe, 38334 Saint-Ismier Cedex, France

5 IRISA-INRIA, Campus de Beaulieu, 35402 Rennes Cedex, France

6 DRDC/BIM, CEA-Grenoble, 17 Avenue des martyrs, 38054 Grenoble Cedex 9, France

7 Serono Genetics Institute, Route Nationale 7, 91030 Evry Cedex, France

For all author emails, please log on.

BMC Bioinformatics 2006, 7:21 doi:10.1186/1471-2105-7-21

Published: 17 January 2006

Abstract

Background

A large variety of biological data can be represented by graphs. These graphs can be constructed from heterogeneous data coming from genomic and post-genomic technologies, but there is still need for tools aiming at exploring and analysing such graphs. This paper describes GenoLink, a software platform for the graphical querying and exploration of graphs.

Results

GenoLink provides a generic framework for representing and querying data graphs. This framework provides a graph data structure, a graph query engine, allowing to retrieve sub-graphs from the entire data graph, and several graphical interfaces to express such queries and to further explore their results. A query consists in a graph pattern with constraints attached to the vertices and edges. A query result is the set of all sub-graphs of the entire data graph that are isomorphic to the pattern and satisfy the constraints. The graph data structure does not rely upon any particular data model but can dynamically accommodate for any user-supplied data model. However, for genomic and post-genomic applications, we provide a default data model and several parsers for the most popular data sources. GenoLink does not require any programming skill since all operations on graphs and the analysis of the results can be carried out graphically through several dedicated graphical interfaces.

Conclusion

GenoLink is a generic and interactive tool allowing biologists to graphically explore various sources of information. GenoLink is distributed either as a standalone application or as a component of the Genostar/Iogma platform. Both distributions are free for academic research and teaching purposes and can be requested at academy@genostar.com. A commercial licence form can be obtained for profit company at info@genostar.com. See also http://www.genostar.org webcite.