[R] PKPD modelling in R

[Ben Bolker]

Marc Schwartz <marc_schwartz <at> me.com> writes:

> 
> On Oct 8, 2012, at 10:04 AM, "Kozielska-Reid MA, Magdalena"
<m.a.kozielska-reid <at> pl.hanze.nl> wrote:
> 
> > 
> > Dear all,
> > 
> > I have two questions:
> > 
> > 1. Does anyone know if nlmeODE package will work once
>  odesolve is discontinued? If i understand well
> nlmeODE depends on odesolve.
> > 2. Can anyone suggest another package I could use to fit
>  complex PKPD models (described using
> differential equations) to data?
> > 
> > Thanks a lot!
> > 
> > Magdalena

 
> You should probably contact the author of nlmeODE to see if he plans
> to update his package to perhaps use deSolve instead as well as
> other CRAN required updates (eg. a NAMESPACE). nlmeODE has not been
> updated in well over two years, which would logically raise some
> suspicions. Presumably nlmeODE will begin to more materially fail
> CRAN checks once odesolve is formally deprecated and may itself be
> deprecated/orphaned as a result. It already has NOTE entries in the
> CRAN checks for problems already found.

 
> If you have not, you might want to look at the CRAN task view here:
> 
>   http://cran.r-project.org/web/views/Pharmacokinetics.html
> 
> which might provide some possible alternatives.

 The listed alternatives are PKfit (also depends on odesolve) and
PKtools (which doesn't appear to do generic ODEs).

  Since deSolve gives very nearly perfect drop-in replacements for
the functions in odesolve, it seems as if bringing one or more
of these packages up to standard shouldn't be hard at all for
someone with the appropriate skills -- but it would a require a 
few hours of that person's time ...

  Ben Bolker