Dynamic simulations of pathways downstream of ERBB-family: Exploration of parameter space and effects of its variation on network behavior

The signaling-network immediately downstream of the ErbB-family is very important in BC and other cancers, especially considering treatment of the excess of function of dominant onco-proteins with onco-protein inhibitors. We studied and implemented dynamic simulations of four downstream pathways. The fragment of the signaling-network we evaluated was described as a Molecular Interaction Map. Our simulations involved 242 modified species and complexes, 279 reversible reactions, 110 catalytic activities. We used Ordinary Differential Equations for our simulations. We started an analysis of sensitivity / robustness of our network, and we systematically introduced fluctuations of total concentrations of independent molecular species. We adopted mostly the strategy of a random sampling of 1000 cases for each instance of increasing numbers of perturbations. Only a small minority of cases showed an important sensitivity, the number of sensitive cases increased moderately for increasing numbers of perturbations. In most cases the effect of introducing virtual mutations and virtual onco-protein inhibitors was more important than the effect of randomly introduced perturbations, this suggests an acceptable robustness of our network. The importance of our work is primarily related to the fact that the complexity of the 39 basic species signaling-network region we analyzed is of difficult intuitive understanding for a "naked" human mind. Dynamic network simulations appear to be an useful support for an "a posteriori" mental comprehension by a cancer researcher of the behavior of a network of this degree of complexity. The present report suggests the feasibility of a computational approach even in the presence of a multiple number of uncertainties about parameter values.