A complementary approach that we have developed is to treat multistimulus or time point data as a coexpression network and then use the topology of the network to identify points of constriction, or bottlenecks. Bottlenecks are predicted to represent points of control for transitions between system states that are important to the underlying conditions being studied. Though the term bottleneck is used in various ways we here define a functional bottleneck to be a gene whose inactivation causes a measurable effect in the expression of downstream targets, acting either directly or indirectly. Identification of and validation of functional bottlenecks predicted by network analysis should provide insight into the LY2109761 TGF-beta inhibitor dynamics of the disease-relevant biological processes and their regulation, and potentially serve as targets for clinical intervention. Neuroprotection against stroke can be induced by preconditioning with Toll-like receptor ligands that activate the innate immune system prior to stroke. Preconditioning with systemic administrations of the TLR4 agonist lipopolysaccharide or the TLR9 agonist CpG-oligonucleotide provides robust neuroprotection against stroke in mice and nonhuman primates. The responses produced by TLR activation depends on many factors such as the TLR ligand, the cell type, and the environment and these responses set off complex signaling cascade that ultimately affect other cell types and systems. Genomic analysis of the response to preconditioning with LPS, CpG-ODN, or brief ischemia shows that TLR signaling pathway is highly regulated. To identify functional bottlenecks with potential roles in TLRmediated responses and neuroprotection, we have gathered temporal high-throughput transcriptomic responses in the brain and blood using microarrays that simultaneously evaluate the expression of,40,000 genes. By analyzing these large datasets together, it is possible to identify genes of regulatory importance TLR signaling in the system that may be missed by examining a single dataset individually. Additionally, inferred networks provide an abstraction of the system in terms of functional modules that are active at different times and/or under different conditions, which allows placement of bottlenecks in the context of the functional dynamics of the system. Previously several studies have used computational and experimental approaches to define the regulatory structure of immune cells responding to TLR stimulus and to identify important players in these systems. We have used inferred networks to characterize macrophage response to TLR agonists and neuroprotection in a stroke model. Ramsey, et al. used a large set of microarray experiments and bioinformatics approaches to define functional modules and the regulatory structure of macrophage response to TLR agonists. Amit, et al. used a microarray experiments followed by high-throughput siRNA perturbation of a large panel of regulators to define a regulatory network in dendritic cells. Finally, Calvano, et al. constructed networks based on the effect of LPS stimulation on leukocytes from human patients. These networks were based on existing knowledge of protein-protein interactions and regulatory relationships and the authors used these networks to identify important subnetworks using differential expression overlaid on the network.