Latest posts

Celebrating 15.000 visitors by going open source!

In 2019 alone, PMXSolutions.com has been visited by over 15.000 visitors! In order to celebrate this overwhelming interest in the website, I have made the Shiny application for pharmacokinetic simulations open source on GitHub! Visit the GitHub repository https://github.com/michielve/PMX_Simulations Collaborate on code GitHub is a great platform to share and collaborate on coding projects. Feel free to use and improve …

Modelling pulsatile profiles in NONMEM

A modellers challenge: getting data in front of you that don’t seem to fit any of the standard effect models and PKPD relationships. Especially in endocrinology, hormonal profiles are rarely in steady-state and vary constantly over time, complicating the application of NLME models. In this post I will discuss 3 degrees of pulsatile profiles on 3 model hormones: Circadian rhythm …

Building your first PBPK model:the basics

In previous posts we referred exclusively to modelling using the top-down, population approach. However, in recent years, physiology-based, bottom-up approaches are getting more attention from both industry and regulators. Population and physiology-based approaches share some common ground: both approaches describe the body as a system of compartments connected by rates to describe drug disposition. If, population approaches use 1, 2 …

Plotting PK/PD hysteresis with variability in R using ggplot

The identification of hysteresis in a PK/PD relationship provides information on a possible delay between the plasma concentration and the effect. The identification of hysteresis can further assist us in structural PK/PD model development by the inclusion of an effect compartment. More information on hysteresis can be found on the following pages: https://www.certara.com/2011/10/26/what-is-hysteresis-in-pkpd-analysis/? https://www.ncbi.nlm.nih.gov/pubmed/24735761 In this post, I will show …

Modelling asymmetry in concentration-effect relationships

Introduction Why do we assume symmetry in our concentration-effect relationship? I recently came across an article published by Piet Hein van der Graaf and Rik Schoemaker from 1999 on the use of asymmetry in concentration-effect curves. “Analysis of asymmetry of agonist concentration–effect curves”: https://www.sciencedirect.com/science/article/pii/S105687199900026X They make a valid point on an assumption we commonly do not investigate: a logistic or …

Validate your model with NPDE analysis

As part of the series of tutorials on model validation, I will get you started on doing your own Normalized Prediction Distribution Errors (NPDE) analysis.  As the VPC and bootstrap, the NPDE is also a simulation-based evaluation tool. NPDEs are useful to investigate the accuracy of the model predictions, and I find it particularly useful when working with models based …

Modelling & Simulation at ASCPT2019 – a short impression

The annual meetings of the American Society for Clinical Pharmacology & Therapeutics have been for me the most inspiring conferences that I visited. This year’s meeting did not disappoint either, here is why. I feel that the PAGE meeting in Europe discusses modelling and simulation from a more technical perspective, whereas the perspective of the ASCPT annual meeting is more …

Get started to non-parametric bootstraps: execution and interpretation

Once you completed your model development, you now have a final model that fits your data the best and you’re now ready to validate your results. One way to internally validate your results is by looking at the precision of your parameter estimates by performing a non-parametric bootstrap. What is a bootstrap? In the figure below I tried to illustrate …

Simulating your pharmacometric model with parameter uncertainty in R

How certain are we about the parameters that we estimate in our population model? Is that volume of distribution that NONMEM gave us really 10 liters with a clearance of 5 L/h? Or can it also be 9L and 6L/h respectively? And what will the impact of this uncertainty be on your simulations? This is the issue that we call …

Creating a simple pharmacometric Shiny application with mrgsolve in R – Part 2

Shiny applications are a great way to show your complicated models in an interactive way. In recent years, many different examples have been published online showcasing a wide range of functionality. Shiny applications are also being increasingly used in pharmacometrics, to quickly see the effect of a different dosing regimen or even develop a Bayesian TDM application. Getting started If …

Creating a simple pharmacometric Shiny application with mrgsolve in R – Part 1

This two part series will show you how to create a simple pharmacometric Shiny application using the mrgsolve package in R. I think that Shiny applications are the most promising way to communicate our complicated mathematical models to others and I would therefore like to show how to create such an application from scratch, including all the codes used with an …