Parameter discovery in stochastic biological models using simulated annealing and statistical model checking.

BackgroundIdentifying a proper model structure, using methods that address both structural and parameter uncertainty, is a crucial problem within the systems approach to biology. And yet, it has a marginal presence in the recent literature. While many existing approaches integrate methods for simulation and parameter estimation of a single model to address parameter uncertainty, only few of them address structural uncertainty at the same time. The methods for handling structure uncertainty often oversimplify the problem by allowing the human modeler to explicitly enumerate a relatively small number of alternative model structures. On the other hand, process-based modeling methods provide flexible modular formalisms for specifying large classes of plausible model structures, but their scope is limited to deterministic models. Here, we aim at extending the scope of process-based modeling methods to inductively learn stochastic models from knowledge and data.ResultsWe combine the flexibility of process-based modeling in terms of addressing structural uncertainty with the benefits of stochastic modeling. The proposed method combines search trough the space of plausible model structures, the parsimony principle and parameter estimation to identify a model with optimal structure and parameters. We illustrate the utility of the proposed method on four stochastic modeling tasks in two domains: gene regulatory networks and epidemiology. Within the first domain, using synthetically generated data, the method successfully recovers the structure and parameters of known regulatory networks from simulations. In the epidemiology domain, the method successfully reconstructs previously established models of epidemic outbreaks from real, sparse and noisy measurement data.ConclusionsThe method represents a unified approach to modeling dynamical systems that allows for flexible formalization of the space of candidate model structures, deterministic and stochastic interpretation of model dynamics, and automated induction of model structure and parameters from data. The method is able to reconstruct models of dynamical systems from synthetic and real data.

A better estimation of surface reaction efficiency of semiconductor-grade graphite with atomic nitrogen, as well as the calibration error are calculated using Bayesian updating based on experimental data. Compared with a conventional deterministic model, the stochastic model approach is a powerful tool in the sense that the model is capable of taking into account underlying error correlations among the data quantities. In this paper, we investigate four different stochastic models (called “stochastic system model classes” herein) corresponding to different descriptions of modeling and measurement error structures, given one deterministic physical model. These stochastic system model classes differ in the covariance matrix structure that is used in the uncertainty model to represent uncertainties associated with the physical model and experimental measurements. For each model class, Bayesian inference is used to estimate the posterior probabilities of the physical model parameters as well as of the stochastic model parameters. Model comparison and selection are then applied based on two measures including Bayesian evidence and Bayesian information criterion, as well as the deviance information criterion. Both measures suggest the stochastic model class, which considers that a correlation between errors in two data quantities among different data points is the most plausible. With the stochastic model class, the range of uncertainty in surface reaction efficiency is estimated to be about two orders of magnitude at .

The aim of the study is to condition stochastic generated realizations on well test data in order to improve simulation of fades and petrophysics in fluvial reservoirs. First we have used the pressure data to estimate the shortest distance from the well to a possible channel boundary and thereby simulate the channel structures. The well test also provides the permeability average in the part of the channel intersected by the well. Together with core/log data and general knowledge of the reservoir this have been used to simulate permeability. These permeability realizations is input to a numerical flow simulator and compared with experimental results of the well test.

A stochastic bivariate ecology model for competing species

Stochastic modeling and simulation of biochemical systems are topics of high interest in Computational Biology. Stochastic mathematical models are critical in accurately capturing the variability observed experimentally in cellular processes, in particular when some species have low molecular numbers. Many, realistic biochemical networks exhibit stiffness, due to the presence of multiple time-scales. For such networks explicit simulation methods are computationally quite intensive. In this thesis, we introduce an improved implicit tau-leaping strategy for the simulation of stochastic biochemical kinetic models. Numerical tests on various biochemical systems of interest in applications show the efficiency of our method.

The structure of oilseed rape is complex due to the development patterns of inflorescence: initiate acropetally but expand basipetally. This unique physiological characteristic makes a big difference in the morphological structure of individual plants in oilseed rape. The time lag between the appearance and elongation of a branch, called “growth delay”, needs to be quantified in order to simulate the developmental pattern of oilseed rape. This study computes the time interval between the appearance and expansion of branches by using a growth delay function. Moreover, a stochastic probabilistic model is built to simulate the differences of the numbers of branches, phytomers of the stems and branches within and among plants for four cultivars. Furthermore, model parameters can be estimated using the experimental measurements for four cultivars (ZY18, ZY50, ZS72 and ZS11) with the generalized least square method. The results indicate that the differences of structure among individual plants for different cultivars can be simulated using the stochastic model. This work provides a new and simple method to describe the differences of structure among plants using a stochastic model. This method can simulate the branch development of oilseed rape, which can be combined to a functional-structural model to simulate the growth and development of oilseed rape.

The uncertainty in the iron loss properties of electrical steel makes traditional deterministic analysis methods insufficient for optimal design and reliability assurance of transformers. In this paper, a stochastic modeling approach for transformer iron loss based on vine copula theory is proposed. A total of 175 electrical steel samples from the same production chain have been investigated. First, the parameters of the deterministic iron loss model are identified and validated using the least squares method. Next, the marginal probability distributions of these parameters are determined and combined with the optimal copula function to model their variability and correlations. A joint probability density distribution has been established using vine copula theory, accurately capturing parameter correlations. The results show that the proposed stochastic model can accurately characterize the probabilistic characteristics of electrical steel losses. Furthermore, the vine copula structure constrains parameter combinations, avoiding the generation of unrealistic magnetization curves. Finally, the stochastic material model is applied to analyze core power losses in a power transformer.

For optimal planning of reservoir design and management, long series or many realizations of daily streamflow are required. Stochastic streamflow models can provide such data. However, considering future changes in climate in these models is challenging. The objective of this study is to compare three non-stationary stochastic models with respect to their performance in simulating daily streamflow for current and future climate conditions. This comparison relies on two non-parametric approaches, namely the k-nn Bootstrap and Simulated Annealing optimization as well as a parametric model working in the frequency domain.All models are run under different experiments: (1) with observed climate from the German Weather Service for a reference and pseudo-future period and (2) with future climate simulations using data from climate models. The simulations from the three models are evaluated for general flow statistics considering current climate and observed changes for a pseudo future. As an additional reference, the HBV rainfall - runoff model driven by observed climate and climate model data is used. The testing of the methods is carried out for some mesoscale catchments in the Harz Mountains comprising streamflow gauges with long daily records. The ability of the stochastic models to simulate the changes for the pseudo future will be the core test for their applicability under changing climate conditions. The results are also expected to demonstrate advantages, disadvantages and limitations of the three methods.

Estimation of Jump-Diffusion Processes via Empirical Characteristic Functions

We consider the problem of verifying stochastic models of biochemical networks against behavioral properties expressed in temporal logic terms. Exact probabilistic verification approaches such as, for example, CSL/PCTL model checking, are undermined by a huge computational demand which rule them out for most real case studies. Less demanding approaches, such as statistical model check-ing, estimate the likelihood that a property is satisfied by sampling executions out of the stochastic model. We propose a methodology for efficiently estimating the likelihood that a LTL property φ holds of a stochastic model of a biochemical network. As with other statistical verification tech-niques, the methodology we propose uses a stochastic simulation algorithm for generating execution samples, however there are three key aspects that improve the efficiency: first, the sample generation is driven by on-the-fly verification of φ which results in optimal overall simulation time. Second, the confidence interval estimation for the probability of φ to hold is based on an efficient variant of the Wilson method which ensures a faster convergence. Third, the whole methodology is designed according to a parallel fashion and a prototype software tool has been implemented that performs the sampling/verification process in parallel over an HPC architecture. 1

Reliability and flexibility are among the key required features of a framework used to model a system. Existing approaches to design resource-constrained, soft-real time systems either provide guarantees for output quality or account for loss in the system, but not both. We propose two independent solutions where each modeling technique has both the above mentioned characteristics. We present a probabilistic analytical framework and a statistical model checking approach to design system-on-chips for low-cost multimedia systems. We apply the modeling techniques to size the output buffer in a video decoder. The results shows that, for our stochastic design metric, the analytical framework upper bounds (and relatively accurate) compare to the statistical model checking technique. Also, we observed significant reduction in resource usage (such as output buffer size) with tolerable loss in output quality.

An industrial site is usually contaminated by accidental (and occasionally intentional) releases of pollutants to the environment from various operations carried out on that site. Consequently, the pattern of contamination created during the life of the site depends in part, at least, on the pattern of operations. Thus, the assessment of the pattern of contamination over the site should be improved: if it is possible to identify the pattern of operations on the site, the duration of the different activities and the perceived likelihood of releases from the different operations. A stochastic model has been developed that can be used to simulate alternative realizations of contaminant releases (duration, extent and timing). The model employs release zones associated with particular activities or groups of activities on the site and the areas of each of the zones may be independent or overlapping. The period of activity in each zone is obtained from the site records, while the likelihood and extent of contamination in each zone is inferred from an analysis of the contamination data obtained by point sampling. The form of the model, the method of inference of the model parameter values from the site data and the application of the model to the study site are presented. The release model has been developed as part of a suite of stochastic models for site ground contamination analysis. The stochastic soil and transport models and the application of the integrated modelling system are described in separate papers.

Piecewise constant (PC) is a stochastic model that can be applied in various fields such as engineering and ecology. The stochastic model contains a noise. The accuracy of the stochastic model in modeling a signal is influenced by the type of noise. This paper aims to propose inverse-gamma noise in the PC model and the procedure for estimating the model parameters. The model parameters are estimated using the Bayes approach. Model parameters have a variable dimension space so that the Bayesian estimator cannot be determined analytically. Therefore, the Bayesian estimator is calculated using the reversible jump Markov Chain Monte Carlo (RJMCMC) algorithm. The performance of the RJMCMC algorithm is validated using data synthesis. The finding is a new PC model in which the noise has an inverse-gamma distribution. In addition, this paper also proposes a parameter estimation procedure for the model based on an RJMCMC. The simulation study shows that the model parameter estimators generated by this algorithm are close to the model parameter values. This paper concludes that inverse gamma noise can be used as an alternative noise in the PC model. The RJMCMC is categorized as a valid algorithm and can estimate the PC model parameters where the noise has an inverse-gamma distribution. The novelty in this paper is the development of a new stochastic model and the procedure for estimating the model parameters. In application, the findings in this paper have the potential to improve the suitability of the stochastic model to the signal.

BackgroundInternational trade regulations require that countries document their livestock's sanitary status in general and freedom from specific infective agents in detail provided that import restrictions should be applied. The latter is generally achieved by large national serological surveys and risk assessments. The paper describes the basic structure and application of a generic stochastic model for risk-based sample size calculation of consecutive national surveys to document freedom from contagious disease agents in livestock.MethodsIn the model, disease spread during the time period between two consecutive surveys was considered, either from undetected infections within the domestic population or from imported infected animals. The @Risk model consists of the domestic spread in-between two national surveys; the infection of domestic herds from animals imported from countries with a sanitary status comparable to Switzerland or lower sanitary status and the summary sheet which summed up the numbers of resulting infected herds of all infection pathways to derive the pre-survey prevalence in the domestic population. Thereof the pre-survey probability of freedom from infection and required survey sample sizes were calculated. A scenario for detection of infected herds by general surveillance was included optionally.ResultsThe model highlights the importance of residual domestic infection spread and characteristics of different import pathways. The sensitivity analysis revealed that number of infected, but undetected domestic herds and the multiplicative between-survey-spread factor were most correlated with the pre-survey probability of freedom from infection and the resulting sample size, respectively. Compared to the deterministic pre-cursor model, the stochastic model was therefore more sensitive to the previous survey's results. Undetected spread of infection in the domestic population between two surveys gained more importance than infection through animals of either import pathway.ConclusionThe model estimated the pre-survey probability of freedom from infection accurately as was shown in the case of infectious bovine rhinotracheitis (IBR). With this model, a generic tool becomes available which can be adapted to changing conditions related to either importing or exporting countries.

Cytoplasmic transport of organelles, nucleic acids and proteins on microtubules is usually bidirectional with dynein and kinesin motors mediating the delivery of cargoes in the cytoplasm. Here we combine live cell microscopy, single virus tracking and trajectory segmentation to systematically identify the parameters of a stochastic computational model of cargo transport by molecular motors on microtubules. The model parameters are identified using an evolutionary optimization algorithm to minimize the Kullback-Leibler divergence between the in silico and the in vivo run length and velocity distributions of the viruses on microtubules. The present stochastic model suggests that bidirectional transport of human adenoviruses can be explained without explicit motor coordination. The model enables the prediction of the number of motors active on the viral cargo during microtubule-dependent motions as well as the number of motor binding sites, with the protein hexon as the binding site for the motors.