Skip to contents

mrgsolve

mrgsolve is an R package for simulation from hierarchical, ordinary differential equation (ODE) based models typically employed in drug development. mrgsolve is free and open-source software.

Resources

Please see mrgsolve.github.io for additional resources, including:

Installation

Install the latest release on CRAN

install.packages("mrgsolve")

Please be sure to see important install-related information here.

Install the current development version

remotes::install_github("metrumresearchgroup/mrgsolve@dev")

Interaction

We welcome questions about anything mrgsolve: installation, getting your model to work, understanding better how mrgsolve works. We also welcome suggestions for how to make mrgsolve more useful to you and to the pharmacometrics community.

Please interact with us at the Issue Tracker. This requires a GitHub account.

Some examples

A simple simulation

Load a model from the internal library

mod <- mread("pk1", modlib())

Simulate a simple regimen

mod %>% 
  ev(amt = 100, ii = 24, addl = 9) %>%
  mrgsim(end = 300, delta = 0.1) %>% 
  plot(CP~time)

A more complicated regimen: 100 mg infusions over 2 hours every 24 hours for one week, followed by 50 mg boluses every 12 hours for 10 days:

mod %>% 
  ev_rx("100 over 2h q 24 x 7 then 50 q 12 x 20") %>%
  mrgsim(end = 600, delta = 0.1) %>% 
  plot(CP~time)

Population simulation 

mod <- mread("popex", modlib()) %>% zero_re()

A data set looking at different patient weights and doses

library(dplyr)

data <- expand.ev(amt = c(100,150), WT = seq(40,140,20)) %>% mutate(dose = amt)

head(data)
.   ID time amt cmt evid WT dose
. 1  1    0 100   1    1 40  100
. 2  2    0 150   1    1 40  150
. 3  3    0 100   1    1 60  100
. 4  4    0 150   1    1 60  150
. 5  5    0 100   1    1 80  100
. 6  6    0 150   1    1 80  150

Simulate

mod %>% 
  data_set(data) %>% 
  carry_out(dose,WT) %>%
  mrgsim(delta = 0.1, end = 72) %>% 
  plot(IPRED~time|factor(dose),scales = "same")

Sensitivity analysis with PBPK model 

mod <- modlib("pbpk")
. Building pbpk ... done.

Reference

  Model file: pbpk.cpp 
  
  $PROB
  # HUMAN PBPK MODEL
  1: Jones H, Rowland-Yeo K. Basic concepts in physiologically based
  pharmacokinetic modeling in drug discovery and development. CPT Pharmacometrics
  Syst Pharmacol. 2013 Aug 14;2:e63. doi: 10.1038/psp.2013.41. PubMed PMID:
  23945604; PubMed Central PMCID: PMC3828005.

Model parameters

param(mod)
. 
.  Model parameters (N=52):
.  name    value  . name  value  . name      value 
.  BP      0.98   | fumic 1      | FVve      0.0514
.  BW      70     | fup   0.681  | HLM_CLint 8     
.  CLrenal 0      | FVad  0.213  | Ka        2.18  
.  CO      108    | FVar  0.0257 | Kpad      0.191 
.  F       1      | FVbo  0.0856 | Kpbo      0.374 
.  FQad    0.05   | FVbr  0.02   | Kpbr      0.606 
.  FQbo    0.05   | FVgu  0.0171 | Kpgu      0.578 
.  FQbr    0.12   | FVhe  0.0047 | Kphe      0.583 
.  FQgu    0.146  | FVki  0.0044 | Kpki      0.597 
.  FQh     0.215  | FVli  0.021  | Kpli      0.57  
.  FQhe    0.04   | FVlu  0.0076 | Kplu      0.62  
.  FQki    0.19   | FVmu  0.4    | Kpmu      0.622 
.  FQlu    1      | FVpl  0.0424 | Kpre      0.6   
.  FQmu    0.17   | FVrb  0.0347 | Kpsk      0.6   
.  FQre    0.104  | FVre  0.0998 | Kpsp      0.591 
.  FQsk    0.05   | FVsk  0.0371 | Kpte      0.6   
.  FQsp    0.0172 | FVsp  0.0026 | .         .     
.  FQte    0.0108 | FVte  0.01   | .         .

Set up a batch to simulate

idata <- expand.idata(Kpli = seq(4,20,2))

idata
.   ID Kpli
. 1  1    4
. 2  2    6
. 3  3    8
. 4  4   10
. 5  5   12
. 6  6   14
. 7  7   16
. 8  8   18
. 9  9   20
mod %>% 
  ev(amt = 150) %>% 
  idata_set(idata) %>%
  mrgsim(end = 6, delta = 0.1) %>%
  plot(Cp~time)