R-based NCA

Case Study

Everything related to Modeling & Simulation and Data Analytics
We make it possible!

R-based NCA

The Challenge

Commercial software Phoenix WinNonlin is the most widely used NCA tool, and as such stands as the main reference tool. The typical use of WinNonlin is via a graphical user interface. Over the last years, a multitude of open-source NCA packages have become available, allowing to script such analyses and to post-process results according to the analyst’s needs. Unfortunately, most of these packages lack certain functionality, user-friendliness, adequate visualization of results, or simply a good documentation and continued development.

The goal for IntiQuan was to develop functionality in R, allowing to efficiently and reproducibly conduct script-based NCA on different types of input data, such as intravenous, bolus, and extravascular, absolute and relative (e.g., weight-based) dosing, and generating the typically required outputs in a tabular and graphical manner that are typically used in reports. Additionally, functionality for efficient reporting of NCA results to Microsoft Word was sought.

The Solution

An agile approach to the development of such an NCA package was chosen. In the first step, the main underlying computational functionality for NCA was implemented, allowing to use an automatic selection of PK samples to consider for the terminal slope calculation but also allowing a modeler to select these points on an individual level. In a second step handling of input datasets was implemented, supporting the modeler in setting up datasets in a correct way. In the third step, functions were implemented to transform NCA results into report ready tabular and graphical representations. Several iterations over these steps were conducted on real NCA projects. An important aspect was to ensure that on one hand the calculation is correct, and at the same time to ensure numerically identical results to the Gold Standard Winnonlin.

The Benefit

NCA can be conducted efficiently within the same environment as modeling analyses, allowing modelers an additional data diagnostic tool. High-quality, report-ready NCA results can be generated in a reproducible script-based manner. Standardized Microsoft Word reports containing the full NCA analysis results can be generated seamlessly and reproducibly without the need for any manual copy-paste approach.

Available as Open Source R package

IQ NCA R package is available as an open source R package from HERE.

 

 

Reproducible Analyses

Computation without Borders and Reproducible Analyses: IQ Desktop

Reporting in Microsoft Word

Conduct of tight-deadline submission projects required the availability of an efficient reporting tool

R based PK/PD modeling workflow

We helped a client leverage the power of open-source modeling tools for internal modeling and simulation tasks