Response to stress is an important biological mechanism to react to environment variations. Different from distinguishing stresses like heat shock, ER stress, and oxidative stress, the study of response to an artificial signal like drug in therapy would be an alternative and also attractive way to understand the cellular response mechanism, which also benefits clinical application. Although differentially expressed genes are usually thought to be therapy responsive genes in many previous researches, more and more attention is diverted from single genes to functions or pathways, in particular for cancer therapy analysis. Thus, comparing with purely molecule (e.g., gene) rewiring, understanding functional reorganization or module rewiring would be more important for systematically studying therapy response or other dynamic biological processes. Therefore, in this paper we propose a model of module network rewiring to characterize functional reorganization, in contrast to gene network rewiring. Specifically, we develop a new framework named as module network rewiring analysis (MNRA) to investigate relevant network modules and their re-connections during an antiviral therapy. In MNRA, we aim to study module dynamics from the network viewpoint, by defining a module network with a module as a node and a path connecting two modules as an edge, which is a network for the molecular interaction system on a higher level. By MNRA experiments on expression data of patients with Hepatitis C virus infection (HCV) receiving Interferon therapy, we found that (1) the consistent module (a set of genes) separates two new subtypes of patients which were not discovered by differentially expressed genes; (2) the patient-group specific module network rewiring reveals necessary functional connections bridged by biological paths; (3) the hierarchical structures of temporal module network rewiring show that they can be taken as spatial-temporal markers to diagnose whether a patient has thera- y response or not. Thus, MNRA indeed can provide biologically systematic clues for potential pharmacogenomic applications and has ability to characterize complex dynamic processes for many biological systems.