Applications of a formal approach to decipher discrete genetic networks

BMC Bioinformatics

Research Article

Laurent Trilling¹  Eric Fanchon¹  Fabien Corblin² 

[1] Laboratoire TIMC-IMAG, UMR CNRS/UJF 5525, Domaine de la Merci, 38710, La Tronche, France;Laboratoire TIMC-IMAG, UMR CNRS/UJF 5525, Domaine de la Merci, 38710, La Tronche, France;Laboratoire IRISA-INRIA centre de Rennes, Campus de Beaulieu, 35042, Rennes, France;

关键词: Boolean Variable;    Boolean Network;    Cellular Context;    Interaction Graph;    Evolution Rule;   

DOI  :  10.1186/1471-2105-11-385

 received in 2010-01-22, accepted in 2010-07-20,  发布年份 2010

来源: Springer

【 摘 要 】

BackgroundA growing demand for tools to assist the building and analysis of biological networks exists in systems biology. We argue that the use of a formal approach is relevant and applicable to address questions raised by biologists about such networks. The behaviour of these systems being complex, it is essential to exploit efficiently every bit of experimental information. In our approach, both the evolution rules and the partial knowledge about the structure and the behaviour of the network are formalized using a common constraint-based language.ResultsIn this article our formal and declarative approach is applied to three biological applications. The software environment that we developed allows to specifically address each application through a new class of biologically relevant queries. We show that we can describe easily and in a formal manner the partial knowledge about a genetic network. Moreover we show that this environment, based on a constraint algorithmic approach, offers a wide variety of functionalities, going beyond simple simulations, such as proof of consistency, model revision, prediction of properties, search for minimal models relatively to specified criteria.ConclusionsThe formal approach proposed here deeply changes the way to proceed in the exploration of genetic and biochemical networks, first by avoiding the usual trial-and-error procedure, and second by placing the emphasis on sets of solutions, rather than a single solution arbitrarily chosen among many others. Last, the constraint approach promotes an integration of model and experimental data in a single framework.

【 授权许可】

CC BY   
© Corblin et al; licensee BioMed Central Ltd. 2010