Uncovering signalling differences between primary and transformed hepatocytes using cell-specific logic-based pathway models

The last years have witnessed remarkable progress in the identification of oncogenes and the signaling networks in which they operate. Pathway maps summarizing literature knowledge are widespread and useful, but they do not allow prediction of pathway operation and do not usually include cell-type specific information. This is a critical limitation because it is precisely these differences between normal and diseased cells that are targeted for therapeutic intervention. We have developed an efficient method to construct cell-specific logic models of signaling networks based on generic pathway maps and high-throughput phosphoproteomics data. These methods are embedded in the MATLAB toolbox CellNetOptimizer, which works in concert with DataRail a complementary toolbox for managing and transforming varied data. We developed models for different cell types that allow us to elucidate significant differences in the wiring of signaling networks. Our results suggest that cancer cell lines are heterogeneous in their signal processing, yet share pathway abnormalities that confer oncogenic properties on them.