Parameter Estimation in Nonlinear Mixed Effect Models Using saemix, an R Implementation of the SAEM Algorithm

Abstract

The saemix package for R provides maximum likelihood estimates of parameters in nonlinear mixed effect models, using a modern and efficient estimation algorithm, the stochastic approximation expectation maximisation (SAEM) algorithm. In the present paper we describe the main features of the package, and apply it to several examples to illustrate its use. Making use of S4 classes and methods to provide user-friendly interaction, this package provides a new estimation tool to the R community.

A Tutorial on RxODE: Simulating Differential Equation Pharmacometric Models in R

“…No licenses are required, and it does not require linking with any external software. Although currently focused on efficient simulations, RxODE can also be used for parameter estimation through the many existing statistical estimation algorithms in R, including nonlinear mixed effects models, stochastic approximation expectation‐maximization (SAEM), and Bayesian methods using Gibbs sampling, e.g., JAGS . Future work includes developing functionality to aid users in linking RxODE models with these estimation algorithms in a more efficient manner.…”

Nonlinear Mixed‐Effects Model Development and Simulation Using nlmixr and Related R Open‐Source Packages

“…A similar need exists for a FOSS NLMEM fitting tool that is compatible with existing workflow and reporting software; can be used in any setting, anywhere, and for any purpose; can be studied, changed, adapted, and improved as needed; and can be redistributed without restriction or licensing fees. Several other projects are ongoing to meet this need, most notably Stan and its extensions PMX Stan and Torsten and the R package saemix …”

“…Several other projects are ongoing to meet this need, most notably Stan 26 and its extensions PMX Stan and Torsten and the R package saemix. 27 nlmixr is a tool for fitting NLMEMs in R. 28 It is being developed as FOSS under the GNU General Public License (GPL) version 2.0, which guarantees end users the freedom to run, study, share, and modify the software as they see fit and stipulates that derivative work can only be distributed under the same license terms or any later GPL license versions (such as GPL-3.0). It is largely built on the previously developed RxODE simulation package for R 29 and has no external dependencies that require licensing.…”

Machine learning and mechanistic modeling for prediction of metastatic relapse in early-stage breast cancer

“…8W8 is the vector of fixedeffects and @ the vector of random effects specific to individual , with standard deviation . Parameters were fitted using lilkelihood maximization implemented in the R package saemix [8]. Once fitted, the model gave personalized distributions of times to relapse in independent patients not used during the training phase, where randomness came from random effects, from which probabilities of 5-years metastatic relapse could be computed.…”