New features and advanced methods for model building
and evaluation in NONMEM VI
(A NONMEM Intermediate Workshop )
The Uppsala Pharmacometrics group is happy to announce a NONMEM Intermediate Workshop on advanced methods for model building and evaluation using NONMEM VI. The Uppsala group has been involved in developing NONMEM models and methodology for over fifteen years. During the last decade they have been involved in the testing and evaluation of NONMEM VI as it has been developed and several research projects have focused on new functionality in NONMEM VI. In parallel, diagnostics for model building have been developed, evaluated and integrated into NONMEM VI-adapted versions of the programs Perl-speaks-NONMEM (PsN) and Xpose 4 (described below). In this course the Uppsala group will present some of the most exciting new features of NONMEM VI and how these tools can be used in the development and evaluation of non-linear mixed effects models.
Course Outline (2.5 days)
The course will last for 2.5 days and consists of both lectures and hands-on computer exercises. Participants will use NONMEM VI, PsN and Xpose 4. PsN and Xpose 4 are freeware and may be installed on participants’ computers for use after the course.
- Day 1 – Model building and advanced numerical diagnostics
- Introduction to new features of NONMEM VI
- Improved model building and diagnostics using PsN and XPOSE 4
- Automation of NONMEM VI runs
- Introduction to diagnostics for NONMEM VI runs
- Case deletion diagnostics
- Typical individual (PRED) based diagnostics
- Hands-on computer exercises with NONMEM VI, PsN and Xpos
- Day 2 – Advanced visual diagnostics
- Shrinkage and empirical Bayes estimates based diagnostics
- Conditional weighted residuals and residual based diagnostics
- Simulation based diagnostics
- Visual and numerical predictive checks
- Hands-on computer exercises with NONMEM VI, PsN and Xpose
- Day 3 – Advanced NONMEM VI features
- The use of prior information on parameter values in NONMEM evaluations
- Using nonparametric parameter estimation in model building using NONMEM VI
- Parameter distribution transformations with estimated shape parameters
- Simultaneous modeling of continuous and categorical data
- Handling data below the limit of quantification (BLQ)
- Hands-on computer exercises with NONMEM VI, PsN and Xpose
Having used NONMEM with an understanding of population PK-PD modeling and/or having attended a NONMEM basic workshop.
The course is to be held at The Westin National Harbor, Maryland, USA (171 Waterfront Street, National Harbor, Maryland 20745, USA).
A limited number of discounted rooms have been booked for this course at the following rates:
NOTE: Group rate available until February 18, 2009. Subject to availability.
Reserve a room at the discounted rates by clicking here.
The course starts the 21st of March 2009 (registration at 1.30 pm, lectures start at 2 pm). The course ends at 5 pm the 23rd of March 2009. There will be a welcome reception on the first day.
Regular fee: 2800 USD
Academic/Government fee: 1800 USD
Registration fee includes: syllabus, continental breakfast, morning and afternoon refreshments, lunch and a welcome reception.
Attendance is strictly limited to 50 with a limited “reduced fee” places available for academic/government participants. Registration and academic fees will be offered on the basis of first come, first served.
Please register online at
PLEASE NOTE: prices on this registration site are in Swedish Kronor, based on the exchange rate for Monday, Feb. 2, 2009.
If you have problems, e-mail Andrew Hooker at email@example.com
Registation is not complete until payment is recieved.
Cancellation: Refunds will be provided on or before December 10, 2008. Registrants are welcome to swap their place with another person at any time.
The course will include hands-on training, with the participants working on their own computers. All programs will run from a USB memory-stick and participants will not be required to install any programs on their computer. NOTE: All participants must bring their own Windows laptops. PsN and Xpose 4 are freeware and may be installed on participants’ computers for use after the course.
Xpose 4 is a model building aid for population analysis using NONMEM. Xpose produces various plots and analyses to facilitate data set checkout, goodness-of-fit analysis, model exploration and visualization, model diagnostics, candidate covariate identification and model comparison. Xpose 4 is written in the programming language R with a modular and object oriented design and allows for integration into nearly any model building platform. New functionality in Xpose 4 allows the user to create simulation based reference diagnostics, create visual and numerical predictive check plots, compute the conditional weighted residuals and modify standard Xpose graphics to create publication quality figures. Xpose 4 also makes use of the new features in NONMEM VI including the visualization of non-parametric parameter distributions and individual objective function contributions. Xpose is freely available at xpose.sf.net.
Perl-speaks-NONMEM (PsN) is a collection of Perl modules and programs aiding in the development of non-linear mixed effect models using NONMEM. The functionality ranges from simpler tasks such as parameter estimate extraction from output files, data file sub setting and resampling, to advanced computer-intensive statistical methods. PsN includes both stand-alone tools as well as development libraries for method developers. PsN is freely available at psn.sf.net.
Prof. Mats Karlsson
Mats Karlsson is professor of Pharmacometrics at Uppsala University, Sweden where he leads a research group of about twenty-five modelers. He received his PhD in pharmacokinetics from this university in 1989 and has been a research fellow at University of Glasgow and University of California, San Francisco, and a visiting professor at Georgetown University, Washington DC. He has received the Giorgio Segre Prize from EUFEPS and is editor for the Journal of Pharmacokinetics and Pharmacodynamics. His research interests focus on methodological aspects of non-linear mixed effects model building and applied PKPD modeling. He has published over one hundred fifty original research articles in the area of PK and PKPD.
Dr. Andrew Hooker
Andrew Hooker is an associate professor of pharmacometrics at Uppsala University, Sweden. Andrew received his PhD in Bioengineering from the University of Washington, Seattle, USA in 2003, and then moved to Sweden as a Pfizer post-doctoral fellow at Uppsala University. His research focuses on methodological problems associated with building and evaluating pharmacometric models as well as optimal experimental design. Andrew is a primary developer of both Xpose 4 and the optimal design program PopED.
Dr. Radojka Savic
Radojka Savic joined the Pharmacometric research group at Uppsala University, Sweden in 2003 after receiving her pharmacist degree from University of Belgrade, Serbia. Since then, she has been actively working on evaluating new features of NONMEM VI. Her research interests focus on nonparametric methods for population analysis, improved diagnostics for NONMEM VI and development of new absorption models.