Simulating different populations

From: "Tsai, Max" max.tsai@spcorp.com Subject: [NMusers] Simulating different populations Date: Wed, 10 Aug 2005 10:35:42 -0400 Can NONMEM perform simulations to incorporate variability associated with fixed effects across populations? When one develops a NM model, NONMEM output lists the final parameter estimates and the standard error estimates for those parameters. Can NONMEM use those standard errors to simulate different populations so that the fixed effects are not so "fixed"? For instance, if SUBPROBS=2, I would like to see one population have one set of fixed effects and a second population have a different set of fixed effects. For a given population, the fixed effects would be determined by the population parameter estimates plus some population variability whose distribution is specified by the standard error. This would be on the population level between replicate studies and thus a step higher than BSV and WSV. Is this capability built into NONMEM? If so, how would it be done? Thanks. -Max Max Tsai, Ph.D. Senior Scientist (DM/PK) Schering-Plough Corporation 2015 Galloping Hill Road K-15-2-2650 Kenilworth, NJ 07033-0530 *: (908) 740-3911 *: (908) 740-2916 *: max.tsai@spcorp.com