A probabilistic model to integrate chip and microarray data

The purpose of this note is to introduce and solve in detail a probabilistic model for integrating microarray data with location data (typically obtained via Chromatine Immunoprecipitation, ChIP). Experimental validation and biological insights obtained using this model will be published elsewhere (Sanguinetti et al. [2006]). The problem we are addressing is a key one in bioinformatics. Cellular processes are assumed to be initiated by the transcription of genes into mRNA and its successive translation into proteins. Transcription is regulated by a complex network of biochemical processes, entailing the binding of transcription factor proteins to the promoter regions of the genes. Given the structure of the network