Network perturbation analysis of gene transcriptional profiles reveals protein targets and mechanism of action of drugs and influenza A viral infection

Noh, Heeju; Shoemaker, Jason E; Gunawan, Rudiyanto

In network-based analysis, one adopts a system-oriented view by using cellular networks, such as gene regulatory network (GRN) and/or protein-protein interaction network (PIN). A number of network-based analytical methods relied on dynamic models of the GRN to infer network perturbations caused by drug treatments (7) (8) (9) (10) (11) . Several notable methods include Network Identification by multiple Regression (NIR) (7) , Mode of action by Network Identification (MNI) (8) , Sparse Simultaneous Equation Model (SSEM) (9) and DeltaNet (10, 11) . In these methods, the GRN is inferred from a training dataset of gene expression profiles using a linear regression derived from a dynamic mechanistic model of the gene transcriptional process. Subsequently, the inferred GRN is utilized for target identification to evaluate deviations in the differential gene expression caused by drug treatments (7) (8) (9) (10) (11) or in disease (12) . A major pitfall of the above methods is that the inference of GRN from gene transcriptional profiles is highly challenging (13) , as the inference problem often becomes underdetermined (i.e. the GRN may not be inferable) (14, 15) . In addition, as mentioned above, the expressions of the drug targets are often unaffected by the drug treatment (3) .