Abstract: The integration of multiple biological interaction networks helps us to identify plausible mechanisms which may involve with a cellular phenomenon under scrutiny. In recent years, biomarkers for predicting adverse reaction of Nevirapine in HIV-infected patients have been reported and successfully applied to screen patients. However, underlying mechanisms that may orchestrate such phenomenon has yet to be identified. This paper presents the development and result of network integration and reconstruction for identifying plausible underlying mechanisms. Human disease network, HIV-1 and human protein interaction network, and protein interaction network were reconstructed from the online database and integrated together to form the baseline interaction network. The biomarkers of Nevirapine adverse reaction of Thai HIV-infected patients were overlaid onto the baseline interaction network. The sub-network that relate to adverse drug reaction was then extracted from the integrated network. This sub-network can be used to identify gene and protein clusters that may associate with the adverse drug reaction. The result suggests that the integration of human disease network and protein interaction network can be effectively used to identify plausible underlying mechanisms of given biomarkers from a genomic study.