Enhances 'xpose' Diagnostics for Pharmacometric Models from 'Certara.RsNLME' and Phoenix NLME

Facilitates the creation of 'xpose' data objects from Nonlinear Mixed Effects (NLME) model outputs produced by 'Certara.RsNLME' or Phoenix NLME. This integration enables users to utilize all 'ggplot2'-based plotting functions available in 'xpose' for thorough model diagnostics and data visualization. Additionally, the package introduces specialized plotting functions tailored for covariate model evaluation, extending the analytical capabilities beyond those offered by 'xpose' alone.

Certara.Xpose.NLME

Overview

Certara.Xpose.NLME is an R package used to creates xpose databases (xpose_data) for PML/NLME results. Additionally, Certara.Xpose.NLME offers various covariate model diagnostic functions, not available in the xpose package.

Installation

Windows

install.packages("Certara.Xpose.NLME",
                 repos = c("https://certara.jfrog.io/artifactory/certara-cran-release-public/",
                           "https://cloud.r-project.org"),
                 method = "libcurl")

Linux

install.packages("Certara.Xpose.NLME",
                 repos = c("https://certara.jfrog.io/artifactory/certara-cran-release-public/",
                           "https://cloud.r-project.org"))

Features

• Create an xpose_data object from NLME model execution directory/files, compatible with xpose functions
• Includes model diagnostic functions for categorical and continuous covariates
• Includes functions to extract estimates from model output into a data.frame for table formatting

Usage

library(Certara.RsNLME)
library(Certara.Xpose.NLME)
library(xpose)
library(ggplot2)


# Create two-compartment pharmacokinetic model
model <- pkmodel(numCompartments = 2, data = pkData, 
                ID = "Subject", Time = "Act_Time", A1 = "Amount", CObs = "Conc",
                modelName = "TwCpt_IVBolus_FOCE_ELS")

# Update parameters and residual error
model <- model |>
  structuralParameter(paramName = "V2", hasRandomEffect = FALSE) |>
  fixedEffect(effect = c("tvV", "tvCl", "tvV2", "tvCl2"), value = c(15, 5, 40, 15)) |>
  randomEffect(effect = c("nV", "nCl", "nCl2"), value = rep(0.1, 3)) |>
  residualError(predName = "C", SD = 0.2)

# Fit model               
job <- fitmodel(model)

# Create xpose_data object from model + fit objects in R using xposeNlmeModel()
xp <- xposeNlmeModel(model, job)

# Alternatively, create xpose_data object from model output directory using xposeNlme()
xp <- xposeNlme(dir = "./TwCpt_IVBolus_FOCE_ELS", modelName = "TwCpt_IVBolus_FOCE_ELS")

# Use xpose plotting function with theme from ggplot2
res_vs_idv(xpdb) + 
  theme_classic()