Research On The Application Of EnKF Algorithm And MCMC Algorithm In Virology

In recent years,virology has become the focus of research the field of biology,which has attracted the attention of many scientists.Thanks to the innovation of biotechnology,our understanding of viruses has also undergone fundamental changes,from the initial understanding of the morphology and basic characteristics of biological cells to the analysis of genomics and bioinformatics.In this paper,different algorithms were used to study the parameter estimation in phage gene regulation network model and plant pathogen model by means of statistics.In chapter 2,we construct a model based on the gene regulatory network of lambda phage and analyze its dynamic behavior.Meanwhile,the Ensemble Kalman Filter(En KF)algorithm was used to compute the ratio of chemical reaction rate constant in the phage gene regulation network and combining the regulatory mechanism of the network to infer the amount of protein to avoid the high cost of protein detection.Simulation results show that the parameter estimation results of the En KF algorithm are better than those of the least squares parameter estimation.Aiming at the problem that the parameter dimension in the plant pathogen model is too high to sample directly,we propose and discuss the application of Markov chain Monte Carlo in the hierarchical Bayesian model in chapter 3.Combining the hierarchical model and the Bayesian method,the M-H algorithm and hybrid MCMC algorithm are used to sample each parameter in the model.Comparing the numerical simulation results of the algorithms,we find that the hybrid MCMC algorithm can better estimate the parameters in the model.This algorithm just avoids the disadvantage that the relatively complex posterior distribution of the individual parameter in the parameter set is difficult to sample.