A Simulation-based comparison between parametric and nonparametric estimation methods in PBPK Models

Abstract

Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. AbstractWe compare parametric and nonparametric estimation methods in the context of PBPK modeling using simulation studies. We implement a Monte Carlo Markov Chain simulation technique in the parametric method, and a functional analytical approach to estimate the probability distribution function directly in the nonparametric method. The simulation results suggest an advantage for the parametric method when the underlying model can capture the true population distribution. On the other hand, our calculations demonstrate some advantages for a nonparametric approach since it is a more cautious (and hence safer) way to assess the distribution when one does not have sufficient knowledge to assume a population distribution form or parametrization. The parametric approach has obvious advantages when one has significant a priori information on the distributions sought, although when used in the nonparametric method, prior information can also significantly facilitate estimation.

A hierarchical Bayesian approach for parameter estimation in HIV models

“…For example, they could either produce a model that fits the data poorly or a reasonable model fit to the data with an incorrect distribution (e.g., see the examples in [8]). In this case, a nonparametric estimation method such as one based on the Prohorov metric framework developed and discussed in [5,6,8] could be more appropriate since it does not require one to provide a distributional structure for the estimated parameters. One such method has been successfully used in PBPK modelling efforts (see [7,8]).…”

“…In this case, a nonparametric estimation method such as one based on the Prohorov metric framework developed and discussed in [5,6,8] could be more appropriate since it does not require one to provide a distributional structure for the estimated parameters. One such method has been successfully used in PBPK modelling efforts (see [7,8]). As will be detailed shortly in a publication, this approach is currently being investigated by our group with the model (1.1) and is yielding quite encouraging results with respect to significant reduction of the computational burden as compared to that for an MCMC approach.…”

“…Before proceeding with our presentation, we note that in our considerations here the methodology investigated is parametric in nature (i.e., we assume functional parameterized forms for various prior distributions) as opposed to the nonparametric approaches developed and discussed in [4][5][6][7][8].…”