We have developed a machine learning framework for the inference of regulatory networks from microarray and proteomic data. It is implemented in the R statistical language, is parallelized and runs efficiently on a multi-core Linux or Mac OS X computer or cluster. The two components of this framework, available as open source packages, are:
1. cMonkey learns context-specific (condition-dependent) modules of co-regulated genes by integrating (a) gene expression data, (b) de novo detection of cis-regulatory DNA motifs, and (c) connectivity in functional association or physical interaction networks.
2. Inferelator identifies the most probable regulatory influencers (environmental factors and/or transcriptional regulators) of each bicluster and can be used to make dynamical predictions of bicluster responses under novel experiments.