Re: ETA on ALAG1

Dear Nidal, Dear All, It is good that there are various options in NONMEM to resolve the difficulty of estimating variability in lag-time. I think there are two additional points: 1) Do we believe that there truly is a notable variability in lag-time? 2) Is it really important to precisely describe the absorption process for the drug / task of interest? Or: Do the other parameter estimates change notably, if the absorption process is in part misspecified? >From my experience: For 1), to the best of my memory, all frequently sampled PK studies I ever modeled benefitted from inclusion of lag time and associated BSV. I would consider gastrointestinal transit as one of the more variable processes in PK, so inclusion of BSV is justified. For 2), Sometimes, estimates for CL and V1 differ between more sophisticated and basic absorption models. (Rada Savic et al. have generated hard data on this). If CL is not affected and one only likes to predict AUC, then it may not be worth the effort to go for one of the advanced absorption models. If predicting Cmax is of interest, then describing the absorption kinetics well is more critical. I have not systematically studied the following. However, my guess is that programs that calculate gradients (like NONMEM) might have more difficulties with estimating variability in lag-time and duration of infusion. My prediction is that MC-PEM, SAEM, MCMC and NPAG are more robust in estimating BSV on such discontinuous processes. (Of course, more hard data are needed on the latter prediction). Kind regards Juergen ----------------------------------------------- Juergen Bulitta, PhD Pharmacometrics, University at Buffalo, NY, USA Phone: +1 716 645 2855 ext. 281, [email protected] ----------------------------------------------- -----Ursprüngliche Nachricht----- Von: <[email protected]> Gesendet: 12.12.08 22:56:01 An: [email protected] Betreff: Re: [NMusers] ETA on ALAG1 As Nick said you could have a negative objective function that is not an error. Usually estimating an ETA on lag time could be problematic; however you could use the hybrid method in NM. Please see the NM help files for more details. Nidal AL-Huniti, PhD Associate Director, Modeling and Simulations ICON Development Solutions On 12/12/08, *Nick Holford* <[email protected]> wrote:Ayyappa, It is not an error to have a negative objective function value. It is an error not to describe the problem accurately. If you want to get help from nmusers then please quote the EXACT wording of the error message and describe what you did that caused the error to appear. Nick [email protected] wrote: Dear All, It is an old topic, but definetely not completely resolved. Can anybody provide tips on estimating variability on ALAG1 in a 2 comp, oral absorption model (I am using NONMEM V). I am getting an error message OBJFUN is negative. I have searched in usernet archive but could not solve the problem. I appreciate your help. Regards, Ayyappa Chaturvedula GlaxoSmithKline 1500 Littleton Road, Parsippany, NJ 07054 Ph:9738892200 -- Nick Holford, Dept Pharmacology & Clinical Pharmacology University of Auckland, 85 Park Rd, Private Bag 92019, Auckland, New Zealand [email protected] tel:+64(9)923-6730 fax:+64(9)373-7090 http://www.fmhs.auckland.ac.nz/sms/pharmacology/holford