Network controllability integrates control theory concepts with complex networks in an attempt to identify optimal control strategies in disparate systems from natural to engineering fields. However, it is unclear what control means in most biological systems. To address this limitation, we use viral infection as a paradigm to model control in an infected cell using available HIV-1 host-interaction data in the context of the human molecular interactome comparing two dynamic states: normal and infected (Fig. 1A). Our network controllability analysis demonstrates how a virus effectively brings the dynamic host system into its control by mostly targeting critical control nodes that are also biological significant in different pathways. This viral infection strategy provides a unique benchmark to discriminate between available network- control models (Fig. 1B) and demonstrates that the minimum dominating set (MDS) method better accounts for how critical and indispensable host proteins are exploited than the maximum matching (MM) method (Fig. 1C). The performance of MDS, that accounts for the inherent non-linearity of signaling pathways, validates for the first time the applicability of the control theory framework for investigating viral use of host cells and gives a more complete and dyna
