1 reply to this topic

[cmukashyaka]

Hello,

I am working on an indirect response model (limited inhibition of Kin), I have baseline values for each subject measured at the start of study. I need to use these values and fix them as E0 where E0 = Kin/Kout so that I can only estimate either Kin or Kout but not both at the same time.

I think there is a way to have these baseline values in a separate column in dataset and set them as covariate in the model code text. Can someone please help me with an example PML code that I can use to accomplish this? 

For dataset, do I add these at each timepoint for each individual as I would for other covariates or only at time=0 where they are actually measured?

 

If my thinking of approaching this is off, how can I take care of this? I don't want to fix E0 to a single value without implementing the interindividual variability observed with baseline measurements. 

 

Thanks a lot

[bwendt@certara.com]

Hi

 

what you are asking for is a typical use of a covariate. You can either use the graphical interface to specify the covariate:

 

 

Or you can specify in textual form. Here is a very generic PML code that you can use as a start. I have entered the baseline values as E0 and used it for definition of Kin, so only Kout will be estimated:

 

test(){

deriv(A1 = - Cl * C)

deriv(E = Kin * (1 - Imax * C / (C + IC50)) - Kout * E)

dosepoint(A1)

C = A1 / V

sequence{E = E0}

E0 = Baseline

error(CEps = 1)

observe(CObs = C + CEps)

error(EEps = 1)

observe(EObs = E + EEps)

stparm(V = tvV)

stparm(Cl = tvCl)

stparm(Kin = E0/tvKout)

stparm(Kout = tvKout)

stparm(Imax = tvImax)

stparm(IC50 = tvIC50)

fcovariate(Baseline)

fixef(tvV = c(, 1, ))

fixef(tvCl = c(, 1, ))

#fixef(tvKin = c(, 1, ))

fixef(tvKout = c(, 1, ))

fixef(tvImax = c(, 1, ))

fixef(tvIC50 = c(, 1, ))

}

 

Best,

Bernd