Understanding co-evolved interaction network involved in allosteric regulation of kinase is of fundamental interest. Here, with aspartokinase III (AKIII) from E. coli as a model system the rearrangement of side-chain interactions upon inhibitor binding are identified by comparing amino acid interaction networks for both the R- and T-states of AKIII. Steered molecular dynamics simulation is then applied to study the dynamic conformational change and the energy transduction process, which is followed by identification of co-evolved interaction network involved in the allosteric regulation. To verify the co-evolved allo-network, mutations of AKIII are examined and modulation of the allosteric regulation is demonstrated by site-directed mutagenesis of the key residues. As illustrated, this study proposes a strategy to identify the co-evolved interaction network that drives the allosteric process. The key feature of the strategy is that key residues involved in the energy transfer pathway can be quickly determined by co-evolutionary analysis of the interfacial interactions of motifs.