Nonlinear Mixed Effects Modeling in R

Date: Saturday, April 2, 2011

Registration Fees: $500 non-students; $50 students

Maximum class size: 50

Instructors/Teaching Assistants: Jose Pinheiro

Workshop summary:

This course will provide an overview of the application of linear and nonlinear mixed-effects models in the analysis of grouped data (such as repeated measures and longitudinal data) using the nlme package in R to illustrate the different stages of model fitting. A unified model-building strategy for the different types of mixed-effects models will be presented and applied to the analysis of real datasets from a variety of areas, including pharmacokinetics, clinical trials, and manufacturing. Strong emphasis will be placed on the use of graphical displays at the various phases of the model-building process, starting with exploratory plots of the data and concluding with diagnostic plots to assess the adequacy of a fitted model.

Schedule: Note, specific times are subject to change based on alignment with concurrent ACoP workshops

8:00 – 9:00 am            Introduction

  • motivating examples of linear and non-linear mixed-effects models
  • a sample of R functions and  methods for fitting and analyzing mixed-effects models.

9:00 – 9:30 am            Grouped data

  • Repeated measures data, longitudinal data, growth curve data
  • Multilevel data
  • Data structures and Trellis displays for grouped data in R

9:30 – 10:30                Fitting linear mixed-effects models

  • Brief review of linear models for independent data
  • The linear mixed-effects (LME) model
  • Using the lme function in R
  • Assessing the adequacy of an LME fit
  • Confidence intervals, hypothesis tests, and predictions

10:30 – 10:45 am         Coffee break

10:45 – 12:00 noon     Extending the basic LME model

  • Modeling the random effects covariance structure
  • Variance functions to model unequal within-group variances
  • Serial and spatial within-group correlation structures

12:00 – 1:00 pm          Lunch

1:00 – 1:45 pm            Nonlinear regression models for independent data

  • Advantages and limitations
  • Estimation and inference
  • Fitting nonlinear models in R: the nls function

1:45 – 3:15 pm            Fitting nonlinear mixed-effects models

  • Nonlinear mixed-effects (NLME) models for grouped data
  • The nlme function in R
  • Assessing the adequacy of an NLME fit

3:15 – 3:30 pm            Coffee break

3:30 – 5:00 pm            Extending the NLME model and covariate modeling

  • Variance functions and correlation structures in NLME models
  • Model building with covariates in NLME models with applications to PK data