JBigin: A Java Package for Bayesian Inference of Genetic Interaction Networks from Micro-Array Data

Bayesian network (BN) has been proposed to infer the gene regulatory networks (GNs) based on differential expressions of its members under various physiological conditions. However, its practical use is limited by the lack of data and the overwhelmingly large number of candidate structures to compare. Here we present JBigin as an integrated tool to infer GN. It allows specification of prior beliefs on the expected GN structure and therefore limits the search space and alleviated the need for expression data. JBigin offers an integrated and extensible API library as well as a biologist-friendly GUI. The usefulness of this package was exemplified by an inference task based on 11 expression profiles. With three expected edges specified as belief, JBigin inferred ten edges, out of which eight edges had been reported in literature. JBigin and its documents can be downloaded at http://www. cls.zju. edu. cn/rlibs/software/jbigin. html.