(711h) A Model Reduction Approach for Mechanistic Biochemical Network Modeling

Cell signaling proteins function through their modular structures, which contain site-specific features, such as binding domains and phosphorylation motifs. To mechanistically understand signaling pathway systems, it is important to develop models incorporating these site-specific features of the signaling proteins. Nevertheless, modeling a signaling network system incorporating such molecular details could be computationally challenging. A key challenge is to deal with the complexity of many possible combinations of site-specific binding and transformations of the molecules. This property, known as combinatorial complexity, can lead to a very large number of species and reactions in a network. This challenge has been partly addressed in a recent technique called the rule-based modeling approach (RBM). However, for systems with high degree of combinatorial complexity, the RBM can become computationally prohibitive. In this work, we present a model reduction framework that can systematically reduce the state-space dimension of a biochemical network model with combinatorial complexity. We demonstrate this framework by developing a highly mechanistic rule-based model of insulin growth factor receptor (IGF1R) signaling in HeLa cells. The model accounts for the detailed membrane-proximal events that lead to IGF1R crosslinking and activation by the extracellular insulin growth factor (IGF1), IGF1R phosphorylation, and protein recruitment in phosphorylated IGF1R. The model in its unreduced form represents a large reaction network system that is computationally intractable. By applying our model reduction technique, we show a dramatic reduction in the network size of the model. We use the reduced model to analyze IGF1-induced temporal and steady-state autophosphoryaltion of six tyrosine sites in the intracellular domain of IGF1R. We also analyze receptor recruitment of 16 HeLa cell-specific signaling proteins that directly compete for the six IGF1R autophosphorylation site binding.