Teaching in population pharmacometrics
line decor
    HOME  > COURSES  :: 
line decor


Advanced methods for population model building, evaluation and usage in NONMEM

Lambertville, NJ, USA
March 9-11, 2020


Pharmacometric modeling has become a pillar in model informed drug development (MIDD). With this comes expectations with respect to quality, efficiency, transparency and innovation in the implementation of the modeling and decision-making process. In this course we will present methods that will help make model building of standard problems more efficient and improve the final product. Further, it will give modelers a larger toolset of diagnostics and model components when it comes to development of models for challenging situations. Automated procedures recently developed for PsN & R facilitates a comprehensive assessment of a model and tailored functionality allow command-line transformations of models.

On March 9-11 in Lambertville, NJ, Mats Karlsson and Andrew Hooker will give a 2.5-day course on “Advanced methods for population model building, evaluation and usage in NONMEM”. The course presents strategies for model building and improvement, the latest methods for model evaluation, as well as strategies to consider when utilizing models for model-informed drug development.

The course consists of both lectures and hands-on computer exercises applying the methods discussed. This hands-on material is based on the most recent developments from NONMEM 7.4, Perl-speaks-NONMEM (PsN) and Xpose. Participants get a vast amount of hands-on examples, code, code snippets and lecture material that can be useful on a daily basis.

If you want to learn how to use tools and methods for fast, efficient and comprehensive model building, evaluation and usage, come join us in March!

Topics covered

  • Model building and model components
    • Overall modeling strategies
    • Random effects models (standard and extended)
    • Residual error models (standard and extended)
    • Mixture modeling
    • Handling censored data (e.g. BQL and dropout)
    • Covariate models and model building
    • Estimation methods and settings
  • Model evaluation
    • Prediction- and Residual-based
    • Empirical Bayes Estimate (EBE) and sampling-based diagnostics
    • Simulation and Simulation-Evaluation/Estimation-Based
    • Outlier and influential individual diagnostics
    • Automated evaluations
    • Covariate model focused diagnostics
    • Parameter uncertainty (bootstrap, SIR, COV)
    • Automated model quality assesment (QA)
  • To consider when applying models for informed drug development
    • Bias assessment
    • Power and Type I error
    • Model averaging

Intended course participants

The course is designed for those who have a good working knowledge of pharmacometric analysis with experience in performing NONMEM analyses and/or have attended a NONMEM basic workshop.


Prof. Mats Karlsson
Assoc. Prof. Andrew Hooker

Practical Information

The course will run March 9 - 11 (10:00 am - 5:30 pm on Day 1, 8:30 am -5:30 pm on day 2 and 8:30 am – 3:00 pm on Day 3) at Lambertville Station Inn & Restaurant, 11 Bridge St, Lambertville, NJ, 08530, USA.

Computer hardware/software - 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. All participants must bring their own Windows laptops (for MAC OS and Linux users a virtual windows environment will work).

Course Fee

Regular fee: €1,500
Student fee: €1,000

Registration fee includes: extensive electronic material including lectures, exercises, programs (PsN and Xpose), additional self-study material including hands-on's and solutions. Morning and afternoon refreshments and lunches are included.

Questions and registration

For questions, please e-mail Andrew Hooker.

To register, follow this link: dinkurs.se/ModBuildEvalUse