NONMEM Users Network Archive

Hosted by Cognigen

[NMusers] the first CRAN release of nlmixr

From: Wang, Wenping <wenping.wang_at_novartis.com>
Date: Thu, 9 Nov 2017 12:42:45 +0000

The nlmixr team is very proud to announce the first CRAN release of nlmixr =
(https://cran.r-project.org/web/packages/nlmixr/), an open source package f=
or population PK and PKPD modelling. nlmixr builds on the RxODE package for=
 simulation of nonlinear mixed effect models using ordinary differential eq=
uations, by implementing parameter estimation algorithms like nlme and SAEM=
 (with gnlmm and FOCEi under active development). nlmixr greatly expands th=
e utility of existing packages (like nlme) by providing an efficient and ve=
rsatile way to specify pharmacometric models and dosing scenarios, with rap=
id execution due to compilation to C++.

This new release features a unified user-interface. This user-interface pr=
ovides PKPD modelers a minimalist, intuitive, expressive, domain-specific m=
odeling language to compactly describe the model structure & efficiently ex=
plore the population data at hand. Below is an example of specifying the t=
heophline data by this domain language:

f <- function(){
         ini({
             lCl <- -3 # log Cl (L/hr)
             lVc <- -1 # log Vc (L)
             lKA <- 1 # log Ka (1/hr)
             add.err <- c(0, 0.2, 1)
             eta.Cl ~ 0.1 # BSV Cl
             eta.Vc ~ 0.1 # BSV Vc
             eta.KA ~ 0.1 # BSV Ka
         })
         model({
             ## First parameters are defined in terms of the initial estima=
tes
             ## parameter names.
             Cl <- exp(lCl + eta.Cl)
             Vc <- exp(lVc + eta.Vc)
             KA <- exp(lKA + eta.KA)
             ## Next, the differential equations are defined
             d/dt(depot) = -KA*depot;
             d/dt(centr) = KA*depot-Cl/Vc*centr;
             ## And the concentration is then calculated
             cp = centr / Vc;
             ## Last, nlmixr is told that the plasma concentration follows
             ## an additive error (estimated by the parameter add.err)
             cp ~ add(add.err)
         })
     }

fit <- nlmixr(f, theo_sd, est="saem")
plot(fit)

Get in touch, reach out to us and discuss ways you could contribute, or try=
 it out for yourself, and experience the joy of exploring and modelling pop=
ulation data all within R! The nlmixr team is working hard to make this com=
munity based modeling tool even better. Our goal: Express your model in nlm=
ixr, openly and freely.

Wenping Wang, PhD
Executive Director, Pharmacometrics
Novartis Pharmaceuticals Corporation
One Health Plaza, East Hanover, NJ 07936-1080, USA
Phone +1 862 7789009; Fax +1 9737818265


Received on Thu Nov 09 2017 - 07:42:45 EST

The NONMEM Users Network is maintained by ICON plc. Requests to subscribe to the network should be sent to: nmusers-request_at_iconplc.com. Once subscribed, you may contribute to the discussion by emailing: nmusers_at_globomaxnm.com.