1 reply to this topic

[jpenzens]

Hi All, 

I am trying to implement a parameter-penalized OF, similar to the Prior functionality in NONMEM.  I am building a PBPK model and would like to estimate parameters if possible, but shrink towards prior in-vitro results if the data is not informative.

 

Is there an elegant or built-in way to implement this outside of directly specifying LL() ? 

Thanks,
Justin 

[smouksassi1]

Dear Justin,

Phoenix NLME does not support a frequentist prior ( PRIOR mentioned above) nor does it have a real bayesian engine.

If you are looking for a Prior informed analysis why not have a look at a Bayesian Software?

You can try to write your own penalized likelihood but then you will have very little out of the box  outputs you will have to do everything on your own.

Bests,
Samer