Mathematical Modeling Of Virus Triggered Innate Immune Signaling Network And Predicting Of Protein Complexes

The innate immune system is the first line of defense against invading microbial pathogens and plays a crucial role in protecting the host from virus infections. In recent years, although the biological experimental study in molecular mechanisms of cellular an-tiviral innate immune and the virus-escaped innate immune response made many great progresses, it is quite difficult to elucidate regulation mechanism for antiviral innate im-mune because innate immune exists delicate and complicated pathogen recognition and signal transduction mechanisms. Based on available the biological experimental data, this dissertation attempts to use mathematical modeling and quantitative analysis to reveal complex molecular mechanisms that regulate the signaling networks of antiviral innate immune response from the system level. Its main works are presented as follows:First, based on biological experimental data, we developed a mathematical model of the virus-triggered signaling pathways that lead to induction of type I IFNs and sys-tematically analyzed the mechanisms of the cellular antiviral innate immune responses, including the negative feedback regulation of ISG56and the positive feedback regulation of IFNs. We found that the time between5and48hours after viral infection is vital for the control or elimination of the virus from the host cells and the parameter counter-factual analysis demonstrated that the ISG56induced inhibition of MITA activation is stronger than the ISG56induced inhibition of TBK1activation. The global parameter sensitivity analysis suggests that the positive feedback regulation of IFNs is very impor-tant in the innate antiviral system. Furthermore, the robustness of the innate immune signaling network was demonstrated using a new robustness index. These results can help us understand the mechanisms of the virus-induced innate immune response at a system level and provide instruction for further biological experiments.Then, based on the above work and new biological experimental data, a simplified core feedback loops module which contained four main components was proposed. We analyzed and compared two classes of models, i.e., the deterministic ordinary differential equations (ODEs) and the stochastic models to elucidate the dynamics and stochasticity in type ? IFN signaling pathways. The bifurcation analysis of the ODE model revealed that the positive feedback plays an fundamental role in regulating the bistable switch, which coupled with the negative feedback coordinately tuned the reversible bistable switch underlying the type ? IFN immune response and exhibits appropriate immune response in response to virus infection. Through quantitative comparing the deter-ministic model with stochastic model, we validated the experimental phenomenon that the expression of IFN is stochastic but the expressing of IFN stimulated genes are not. Furthermore, through characterizing the intrinsic noise and extrinsic noise by us-ing corresponding Master and Langevin equations, respectively, the simulation results showed that the intrinsic noises have slight influences on bistable switch but the extrinsic noises characterized by Langevin equations make the distance between two stable states smaller. Moreover, we proposed a multi-state stochastic model based on above deter-ministic model to describe the multi-cellular system coupled by diffusion of IFNs and the corresponding stochastic simulation algorithm to make simulation of the coupled stochastic model. The numerical simulation analysis showed that the positive feedback as well as extrinsic and intrinsic noises has little effects on the stochastic expression of IFNs, but the negative feedback of ISG56on the activation of IRF7has great influence on IFN stochastic expression. Together, these results revealed that positive feedback stabilizes the IFN gene expression and negative feedback may be main contribution to the stochastic expression of IFN gene in the virus-triggered type I IFN response. These findings will provide new insight into molecular mechanisms of virus-triggered type I IFN stochastic expression.Finally, a novel identification method was proposed through assessing the compact-ness of local sub networks by measuring the number of three node cliques. The present method detects each optimal cluster by growing a seed through maximizing the com-pactness function. To demonstrate the efficacy of the new proposed method, we evaluate the performance of the new method using three PPI networks on three reference sets of yeast protein complexes and compare the performance of new method with some of the well known protein complexes detection methods in terms of Fraction of matched complex, Sensitivity, Positive predictive value, Accuracy and Maximum matching ratio. The results show that the protein complexes generated by the new proposed method have better quality than those generated by some previous four classic methods. There-fore, the new proposed method is effective and useful in detecting protein complexes in PPI network. At last, we applied the new algorithm on the human and mouse an-tiviral innate immune signal transduction networks which got by high throughput data, some meaningful results are obtained through the analyze of the predicted the protein complexes.