Experimental evidence suggests that epidermal growth factor receptor (EGFR)-mediated activation of the signaling protein phospholipase Cgamma plays a critical role in a cancer cell's phenotypic decision to either proliferate or to migrate at a given point in time. Here, we present a novel three-dimensional multiscale agent-based model to simulate this cellular decision process in the context of a virtual brain tumor. Each tumor cell is equipped with an EGFR gene-protein interaction network module that also connects to a simplified cell cycle description. The simulation results show that over time proliferative and migratory cell populations not only oscillate but also directly impact the spatio-temporal expansion patterns of the entire cancer system. The percentage change in the concentration of the sub-cellular interaction network's molecular components fluctuates, and, for the 'proliferation-to-migration' switch we find that the phenotype triggering molecular profile to some degree varies as the tumor system grows and the microenvironment changes. We discuss potential implications of these findings for experimental and clinical cancer research.