search

nlmixrdevelopment / nlmixr

nlmixr: an R package for population PKPD modeling
https://nlmixrdevelopment.github.io/nlmixr/
GNU General Public License v2.0
114 stars 45 forks source link

patients with multiple independent treatments #576

Closed fernandoortega closed 2 years ago

fernandoortega commented 2 years ago

Hi,

I want to use nlmixr to determine the parameters of a PD model (I have the PK model). In this dataset, each patient receives several independent treatments. (please see a simplified table below). Would it be possible to use nlmixr to determine PD model parameters and the estimate for each patient?

Thank you very much in advance. Fernando

EXP ID TIME EVID AMT DV CMT 1 1 0 0 0 100 3 1 1 1 101 100 0 1 1 1 2 0 0 80 3 1 1 3 0 0 75 3 2 1 0 0 0 120 3 2 1 1 101 150 0 1 2 1 2 0 0 115 3 2 1 3 0 0 116 3 3 2 0 0 0 100 3 3 2 1 101 25 0 1 3 2 2 0 0 80 3 3 2 3 0 0 75 3 4 2 0 0 0 120 3 4 2 1 101 100 0 1 4 2 2 0 0 115 3 4 2 3 0 0 116 3

mattfidler commented 2 years ago

Hi @fernandoortega

In general the goal of nlmixr is to characterize the population and individual effects. However many models are disease specific, so it is difficult to say without running a full analysis what the best PD model is for your dataset.

Nick Holford (@nholford) has an excellent example of this here:

http://holford.fmhs.auckland.ac.nz/research/nlmixr/

This is what the multiple endpoint model vignette is based on as well:

https://nlmixrdevelopment.github.io/nlmixr/articles/multiple-endpoints.html

I would also suggest looking at xgxr to explore your data. An example is here:

https://nlmixrdevelopment.github.io/nlmixr/articles/xgxr-nlmixr-ggpmx.html