Clustering-based convergence diagnostic for multi-modal identification in parameter estimation of chromatography model with parallel MCMC

2021/07/15 by Zhu, Yue-Chao, Sun, Zhaoxi, He, Qiao-Le
#FOS: Biological sciences #FOS: Mathematics #Numerical Analysis (math.NA) #Quantitative Methods (q-bio.QM)

paper · doi:10.48550/arxiv.2107.07203

Abstract

Uncertainties from experiments and models render multi-modal difficulties in model calibrations. Bayesian inference and mcmc algorithm have been applied to obtain posterior distributions of model parameters upon uncertainty. However, multi-modality leads to difficulty in convergence criterion of parallel mcmc sampling chains. The commonly applied \widehatR diagnostic does not behave well when multiple sampling chains are evolving to different modes. Both partitional and hierarchical clustering methods has been combined to the traditional \widehatR diagnostic to deal with sampling of target distributions that are rough and multi-modal. It is observed that the distributions of binding parameters and pore diffusion of particle parameters are multi-modal. Therefore, the steric mass-action model used to describe ion-exchange effects of the model protein, lysozyme, on the sp Sepharose ff stationary phase might not be fully capable in certain experimental conditions, as model uncertainty from steric mass-action would result in multi-modality.

Related

• Chromatographic Peak Shape from Stochastic Model: Analytic Time-Domain Expression in Terms of Physical Parameters and Conditions under which Heterogeneity Reduces Tailing
• A parallelizable sampling method for parameter inference of large biochemical reaction models
• A nested loop for simultaneous model topology screening, parameters estimation, and identification of the optimal number of experiments: Application to a Simulated Moving Bed unit
• PARAFAC2×N: Coupled Decomposition of Multi-modal Data with Drift in N Modes
• A clustering-based biased Monte Carlo approach to protein titration\n curve prediction