Computational prediction of macrophage specific regulatory network

Brendan Tse1, Timothy Ravasi2, Christine Wells, Yi-Ping Phoebe Chen, David Hume, University of Queensland;, University of Queensland

In the past the larges transcriptional network study on eukaryote had only been carried out on a Saccharomyces cerevisiae model. Studying the underlying regulatory relationship of a complex system with large-scale expression data is often difficult due to large variance in the expression pattern which causes the statistical predictions to be imprecise and ineffective. Our laboratory has collected a large set of microarray information concerning genes that are expressed specifically in cells of the macrophage lineage or inducible by the many signals that regulate macrophage function. With the increased number of regulatory mechanisms accompanied by the recent advance in data collection and analysis, regulatory studies of a well-characterized mammalian system are possible. The completed human and mouse genome sequences with rat to follow in the near future provide a powerful support for identification of conserved mammalian transcription networks. This project aims to create and verify an analytical pipeline system that links existing pattern discovery tools to allow automation of transcriptional elements pattern predictions to be performed simultaneously across multiple species directly from microarray experimental results. The system will be used to map transcriptional pathways to a macrophage specific regulatory network. The use of expressional microarray experiments will allow genes with similar regulatory signals to be clustered together. The system will utilize the unique identifiers from mouse target genes in the microarray experiments. The system aims to link these unique identifiers in the mouse to the promoters of conserved, syntenic genes in other species, and thence to create a database of conserved elements common to any set of co-regulated macrophage expressed gene. Analysis of conserved elements serves to narrow the search space to allow more stringent regulatory pattern prediction.