Bayesian parameter estimation for biochemical reaction networks using region-based adaptive parallel tempering

Abstract

MotivationMathematical models have become standard tools for the investigation of cellular processes and the unraveling of signal processing mechanisms. The parameters of these models are usually derived from the available data using optimization and sampling methods. However, the efficiency of these methods is limited by the properties of the mathematical model, e.g. non-identifiabilities, and the resulting posterior distribution. In particular, multi-modal distributions with long valleys or pronounced tails are difficult to optimize and sample. Thus, the developement or improvement of optimization and sampling methods is subject to ongoing research.ResultsWe suggest a region-based adaptive parallel tempering algorithm which adapts to the problem-specific posterior distributions, i.e. modes and valleys. The algorithm combines several established algorithms to overcome their individual shortcomings and to improve sampling efficiency. We assessed its properties for established benchmark problems and two ordinary differential equation models of biochemical reaction networks. The proposed algorithm outperformed state-of-the-art methods in terms of calculation efficiency and mixing. Since the algorithm does not rely on a specific problem structure, but adapts to the posterior distribution, it is suitable for a variety of model classes.Availability and implementationThe code is available both as Supplementary Material and in a Git repository written in MATLAB.Supplementary information Supplementary data are available at Bioinformatics online.

Reoccurring neural stem cell divisions in the adult zebrafish telencephalon are sufficient for the emergence of aggregated spatiotemporal patterns

“…To calculate the log likelihood across several hemispheres, every single likelihood per hemisphere is summed up and optimized in parallel to form a combined likelihood. We use the PESTO toolbox [ 78 ] to maximize the likelihood including uncertainty estimation via posterior sampling [ 79 ] as can be seen exemplarily in Fig 2B for hemispheres with Δt = 32 h and for all hemispheres in S2C Fig .…”

A multi-data based quantitative model for the neurovascular coupling in the brain

“…In addition, eSS was used in conjunction with two local optimization solvers: dynamic hill climbing [82] and the interior point algorithm included in FMINCON (MATLAB and Optimization Toolbox Release 2017b, The MathWorks, inc., Natick, Massachusetts, United States) paired with the calculation of the objective function gradient using AMICI. For the model uncertainty analysis, we employed the ‘Parameter EStimation ToolBox’ (PESTO) [83] and the region-based adaptive parallel tempering algorithm [84] to generate the posterior distributions of the parameters. The parameter bounds were [-4.5,4.5] in log 10 space unless specified otherwise in Section 2.2.…”

Accelerated regression-based summary statistics for discrete stochastic systems via approximate simulators

“…A large set of methodologies exist for parameter estimation of biochemical systems, especially when the dynamics are modeled using ODEs. Extensive reviews of some of the prominent optimization, metaheuristic, and Bayesian schemes are detailed in [10][11][12][13]. In this work, we are interested in Bayesian parameter estimation when the dynamics of the biochemical system exhibit intrinsic stochasticity and are modeled using a continuous time, discrete-space Markov process [4].…”