4 replies to this topic

[Xiuqing Gao]

Question1: My drug is going through the enterohepatic recycling, but it had second peak at 3hr and third peak at 24hr, how can i build the texual mode with this situation? Can I manually add the another switch variable?

 

Question2: With enterohepatic recycling case, how can I do popPK instead of individual PK in textual model?

 

 

 

 

 

 

Thanks!

[bwendt@certara.com]

Hi Xiuqing

 

re 1)

in the simplest case you would define just another interval:

 

Ri1=3

period=12

Ri2=period+Ri1-Tau

Rate = Switch/Tau

sequence{

Switch=0

sleep(Ri1)

Switch=1

sleep(Tau)

Switch=0

sleep(Ri2)

Switch=1

sleep(Tau)

Switch=0

}

 

 

re 2)
you can easily switch between an individual and a population model
you just need to define random effects:
if you define central volume in individual model the PML code is:
stparm(Vc= tvV)
for a population model you would need to change this line to:
stparm(Vc= tvV*exp(nV))
you can do that for all your structural parameters in similar fashion
now that you have defined random effects with your structural parameters
you would also need to give them initial estimates:
ranef(diag(nV, nCl) = c(0.1, 0.1))

 

Bernd

[Xiuqing Gao]

Thank you, Bernd. One more question: how can I set up covariate stepwise in textual model?

[bwendt@certara.com]

Xiuqing,

 

as discussed through live-chat, please see example project file attached. Here, I have added random effects to the 3-peak model for which I have shown the PML code above. For your information I have added a multiple peak version as well. Hope, this is useful. 

If you think about it, there is yet another source of variability: depending on patient characteristics EHC may happen or not. If you are interested we will present this scenario as part of a new webinar series:

 

https://www.linkedin...514924209643522

 

 EHC_popPK_multPeaks.phxproj   1.18MB   365 downloads

[bwendt@certara.com]

Regarding covariates, I would recommend to use the graphical model builder since it is much simpler, but if you want to do the textual coding, you start by defining the covariates, e.g. 

 

fcovariate(Wt)

fcovariate(Sex())

 

​where Wt is a continuous and Sex a categorical covariate.

 

Now you need to specify covariate effects. If you want to check an effect of Wt on Volume you need to update the structural parameter:

 

stparm(V = tvV * wt^dVdwt * exp(nV))

 

the new covariate effect dVdwt needs to be defined as well:

 

fixef(dVdwt(enable=c(0)) = c(, 0, ))

 

If you want to practice I would recommend you do a few steps in graphical building mode and check the corresponding PML. This will help you to quickly learn the syntax.

 

Bernd