Orientation to Population PK/PD Modeling

Build the conceptual foundation for population PK/PD modeling before moving into data preparation, model fitting, diagnostics, and simulation.
Tip

Module goal: Understand what population PK/PD modeling is, why it is useful, and how a typical modeling workflow is organized.

Module Overview

Population PK/PD modeling combines pharmacology, statistics, mathematics, and clinical interpretation. Before writing nlmixr2 model code, it is important to understand what the model is trying to represent.

In this module, we introduce the core language of population modeling:

  • individuals and populations
  • fixed effects and random effects
  • structural models and statistical models
  • variability and uncertainty
  • PK and PD endpoints
  • model-based simulation
  • the role of nlmixr2 in the modeling workflow

The goal is not to estimate a full model yet. The goal is to build the mental framework that will make the later applied modeling lessons easier to understand.

Learning Objectives

By the end of this module, you will be able to:

  • Explain the difference between individual and population PK analysis.
  • Describe fixed effects, random effects, and residual error.
  • Distinguish structural model components from statistical model components.
  • Explain why mixed-effects models are useful in pharmacometrics.
  • Describe how PK models can be extended to PK/PD models.
  • Recognize the major steps in a population modeling workflow.
  • Understand where nlmixr2 fits into the open-source pharmacometrics ecosystem.

Lessons in This Module

Lesson 1: What Is Population PK/PD Modeling?

This lesson introduces the big picture of population modeling and explains why we need models that describe both typical behavior and individual variability.

Lesson 2: Anatomy of a Population Model

This lesson introduces the major components of a population model: structural model, fixed effects, random effects, residual error, and covariates.

Lesson 3: The Population Modeling Workflow

This lesson walks through the end-to-end modeling process from data preparation to simulation and reporting.

Lesson 4: Orientation to nlmixr2

This lesson introduces the nlmixr2 ecosystem and previews the basic structure of an nlmixr2 model function.

Software Used in This Module

This module is mostly conceptual, but we will occasionally use R for small demonstrations.

library(tidyverse)

Later modules will introduce:

library(nlmixr2)
library(rxode2)
library(nlmixr2data)
library(ggPMX)
library(xpose.nlmixr2)

How This Module Fits the Course

This module prepares you for the rest of the course.

After this module, we will move into:

  • preparing modeling datasets
  • visualizing concentration-time data
  • writing structural PK models
  • fitting models with nlmixr2
  • diagnosing model performance
  • adding covariates
  • simulating dosing and response scenarios

The concepts introduced here will return throughout the course.