To capture the spatiotemporal dynamics of biomolecules for examining the pharmacological action of drugs, we developed a novel protein network analysis, Protein Localization and Modification-based Covariation Network (PLOM-CON) analysis method. This method quantifies the temporal changes in quantity, quality (post-translational modifications such as phosphorylation), and (co)localization of proteins that are activated or inactivated in response to specific signals input from the immunofluorescence images, and visualizes them as a "covariation network" using the "strength of temporal correlation" of the feature quantities. 
In this study, we performed PLOM-CON analysis method to obtain covariation networks for ~50 proteins in insulin-stimulated rat hepatoma H4IIEC3 cells. The results showed that both Akt, a central molecule in insulin signaling, and its phosphorylated form p-Akt (Ser473) are the hub of the network. Furthermore, we revealed that the actin domain, a specific structure temporally created upon insulin stimulus, is the site of accumulation of various molecules for the initiation of glycogen synthesis via phosphorylated GSK3β. Thus, the PLOM-CON analysis method can visualize how signals are widely propagated into cells, and can be used as a technology that can capture the main and side effects of drugs.