Sparse Dynamical Network Reconstruction: the EGFR network case

The ability to reconstruct and identify intracellular protein signaling and biochemical networks is of critical importance in biology today. However, the ability to dynamically measure and collect data from every protein/node within the network is impossible with current methodologies. Consequently, approaches are needed that can use experimentally collected data to accurately reconstruct and extrapolate the higher dimensional network. We sought to develop a mathematical approach to this problem using one of the most well-studied and clinically important signaling networks in biology today, the epidermal growth factor receptor(EGFR) driven signaling cascade. More specifically, we suggest a method for the identification of links among nodes of ordinary differential equation networks from a small set of trajectories with different initial conditions. This method uses specific sparsity arguments that are tailored to the needs of often ill-conditioned systems of representation that arise from the collection of all given trajectories. The enforcement of sparsity allows to consider potentially very large spaces of models and to still be able to detect with high accuracy the few relevant links among nodes, even when moderate noise is added to the measured trajectories. After showing the performance of our method on a model of the EGFR protein network, we sketch briefly the potential future therapeutic applications of this approach.