Simultaneous fit of IV&Oral (FO-Multiplicative)

Dear Simon, I tried fitting simultaneously both IV and Oral data, using naivepooled and FO estimation methods (additive error). However when i use FO with multiplicative error, the standard errors are not computed. Plz find the attached proj. Looking forward for your guidance, With best regards, Raghav, [file name=123_IV_PO.phxproj size=1500818]/extranet/media/kunena/attachments/legacy/files/123_IV_PO.phxproj[/file]

Hi Raghav, sorry for the delay in my response - been rather busy with training courses and of course the 6.2 release. I just opened your project with that release and re-executed your models and I got the SEs without any edits or changes; I have attached the completed project where I also added a Model comparer object, you should be able to open this in 6.1, note if you try to re-run that model you’ll need to re-map ID as a Sort since in 6.2 there is now an explicit ID input. If you do re-run in 6.1 you will see this; and in the Core Status " Warning: Hessian based standard error- computation failed in Cholesky factorization step" Page 73 of the 6.2 release notes will point you to the problem I think you were having in 6.1 and the solution implemented in 6.2; WinNonlin Phoenix Models: when std errors from VIF are used because Hessian failed, need a more prominent warning (QC 7676): Standard errors of estimates are normally computed from the Hessian of the objective function. In the case where the Hessian is not positive definite and the model has Gaussian observations, the standard errors are computed from the variance inflation factors. If this occurs, it will be indicated in the Core Status text output only. Simon. Lastly note we have also added a Log-Additive error model which may be superior for certain problems in this release. [file name=2935323_IV_PO_v62.phxproj size=1887846]/extranet/media/kunena/attachments/legacy/files/2935323_IV_PO_v62.phxproj[/file]2935323_IV_PO_v62.phxproj (1.8 MB)

I was discussing this model with another user and I wanted to mention that Chapter 9 of the NLME Examples guide gives a worked example using the graphical editor of how to obtain this code (for 1 com); test(){ deriv(A1 = - (Cl * C) + (Aa * Ka)) urinecpt(A0 = (Cl * C)) deriv(Aa = - (Aa * Ka)) C = A1 / V dosepoint(A1, idosevar = A1Dose, infdosevar = A1InfDose, infratevar = A1InfRate) error(CEps = 0.1) observe(CObs = C * (1 + CEps)) dosepoint(Aa, idosevar = AaDose, infdosevar = AaInfDose, infratevar = AaInfRate) stparm(V = tvV * exp(nV)) stparm(Cl = tvCl * exp(nCl)) stparm(Ka = tvKa * exp(nKa)) fixef(tvV = c(, 1, )) fixef(tvCl = c(, 0.2, )) fixef(tvKa = c(, 1, )) ranef(diag(nV, nCl, nKa) = c(0.1, 0.1, 0.1)) } Secondly some people using IE as a browser get a message that it’s password encrypted - It’s not- just download it and ensure you have the file extension as “phxproj” and you can open it in PHX fine. Simon

Dear Simon, thanQ for the response, I have another query regarding the Simulation function in population mode. My work flow is as follows: 01. population analysis for plasma concentration profile from rat 02. population analysis for plasma concentration profile from dog 03. scale the final parameters from the above to human Now i have the final parameters in human scaled from animals, can i enter these parameters and simulate for human plasma concentration profiles with uncertaininty. if yes, can be done at steady state also ? With best regards, Raghav