the patients. Previous studies showed that thermoregulatory responses are related to the extubation time. This study had 2 main objectives: (i) to quantify the dynamics of the relation between body (Tbo) and skin (Ts) temperature of coronary bypass patients in response to removal of the heating blanket and (ii) to relate these dynamics with the length of stay (LOS) at intensive care unit (ICU). Methods: For this preliminary study, 7 coronary bypass patients were observed after surgery starting from the moment they arrived at the ICU. During the first 30 minutes at ICU, the patients were warmed by a forced-air heating blanket. The LOS of the patients ranged from 25 to 410 hours. Skin temperature (toe), core body temperature (blood), heart rate (beats/min), medication (mg and mL), and forced-air warming by heating blanket and so on, were measured for a period of 8 hours. The sampling interval of the recorded temperature time series was 1 minute. A data-based transfer function modeling approach was used to determine the dynamics of the temperature profiles. The models described the relation between the blanket temperature (Tbl), the core body temperature (model inputs), and the skin temperature (model output). Results: The developed models were very accurate (R 2 N 0.93). For every patient, second-order models were found representing the relations between the model inputs (Tbo and Tbl) and the model output (Ts). Based on the preliminary experiments, the model results suggested that the feedback gain between Tbo and Ts is related to the length of stay (hours) at ICU. More specifically, patients with a larger LOS showed smaller feedback gain values. Conclusions: To conclude, this study suggests that the dynamics of the thermoregulatory response to surgery of coronary bypass patients can reveal information of the patients' postoperative recovery. http://dx.doi.org/10.1016/j.jcrc.2012.10.028 Abstract 13 Combined network and mathematical modeling suggests novel inflammation positive feedback circuit and role for IL-1 in experimental trauma/hemorrhage and hepatocyte hypoxia Nabil Azhar, Rami Namas, Bahiyyah Jefferson, Derek Barclay, Ruben Zamora, Yoram Vodovotz University of Pittsburgh, Pittsburgh, PA, USA Objectives: Trauma/hemorrhagic shock (T/HS) is the most common cause of death for young people in the United States, costing more than $400 billion annually. Most deaths from T/HS occur due to the multiple organ dysfunction syndrome. Multiple organ dysfunction syndrome is thought to be due, in part, to dysregulated inflammation. We have used computational modeling to define key networks and mechanisms of post-T/HS acute inflammation. We used a Dynamic Bayesian Network (DBN) algorithm (adapted from Grzegorczyk and Husmeier, Bioinfor- matics 2010;27:693). Given time-series data, DBNs provide a way of inferring causal relationships based on probabilistic measures. This approximation helps to suggest the overall network structure, including central nodes that exhibit feedback. Methods: C57Bl/6 mice were subjected to T/HS for 0 to 4 hours. Separately, C57Bl/6 hepatocytes were subjected to an in vitro surrogate of T/HS, namely, hypoxia (culture at 1% O 2 for 1 to 72 hours). Cytokines/chemokines were assayed in the plasma (in vivo T/HS) as well as supernatants and lysates (in vitro hypoxia) by Luminex (MiraiBio, Alameda, CA), and the data were subjected to DBN analysis. Results: Dynamic Bayesian Network analysis suggested a core network that involves the cytokines IL-10 and IL-1α, as well as the chemokines IP-10 and MIG. Assigning biologically plausible functions to directed edges of the network resulted in a so-called incoherent type I feed forward loop that predicted pulsatile, “gatekeeper” behavior for the cytokine IL-1α—a behavior that was indeed observed in preliminary studies in mice subjected to T/ HS. An initial ordinary differential equation (ODE) model was constructed to analyze this behavior in further detail. Simulated trajectories of cytokines and chemokines in the ODE model corresponded closely with the data. Conclusions: Data-driven (DBN) and mechanistic (ODE) analyses suggest a novel inflammatory circuit with a central role for IL-1α in T/HS and hepatocyte hypoxia. http://dx.doi.org/10.1016/j.jcrc.2012.10.029 Abstract 14 Sensitivity of human immune response to influenza a virus infection and its dependence on virus and host phenotypes Sarah Lukens a , Gilles Clermont b , David Swigon a a Department of Mathematics, University of Pittsburgh, Pittsburgh, PA, USA b Department of Critical Care Medicine, University of Pittsburgh, Pittsburgh, PA, USA Objectives: To design a phenotype-based abstraction of host immune response to influenza A infection and its application in an agent-based population-level model of a pandemic. Methods: We developed a novel deterministic differential- equations model of a host-level response to influenza A virus (IAV) in which systemic and respiratory symptoms are mapped to interferon levels and the extent of damage to epithelial cells of the respiratory tract, respectively. We use an ensemble model approach to generate a distribution of parameters that represents the likelihood of the model fitting experimental data consisting of viral titers and symptoms. The resulting ensemble of models reflects patient and virus strain variability, identifiability and sensitivity of the model, and uncertainty in the data. The ensemble is computed using MCMC sampling enhanced by parallel tempering. We use the model to predict the variability of clinical factors, such as onset and severity of symptoms, across the ensemble. We further use sensitivity analysis across the ensemble to relate population-scale clinical phenotypes (severity of infec- tion, immunogenicity) to model parameters. Results: Although marginal distributions of ensemble parameters vary over wide ranges, the sensitivities of trajectories to parameter changes are similar across the ensemble and charac- teristic of the model structure. Using principal component analysis, we compute linear combinations of parameters that represent predominantly virus specific (eg, virulence) or predom- inantly host-specific (eg, immunogenicity) phenotypes. By varying these phenotypes, we obtain a response surface depicting range of biologically relevant host responses to IAV infection. Applying the same methodology across the ensemble yields probabilistic response surfaces that provide a statistical represen- tation of the time course of infection that can be efficiently integrated in a population level model. e7 Abstracts