Complex Bayesian Modeling Workflows Encoding and Execution Made Easy With a Novel WinBUGS Plugin of the Drug Disease...

Area of Science:

• Pharmacometrics
• Computational Biology
• Systems Pharmacology

Background:

• Pharmacometric models are crucial for understanding drug-disease dynamics.
• Interoperability and reproducibility are significant challenges in complex pharmacometric modeling.
• Existing frameworks often lack seamless integration with diverse computational tools.

Purpose of the Study:

• To present the integration of a WinBUGS plugin within the Drug Disease Model Resources (DDMoRe) Interoperability Framework (IOF).
• To demonstrate the IOF's capability in executing pharmacometric models across multiple target tools, including WinBUGS.
• To showcase the application of the IOF in a real-world diabetes study for estimating insulin secretion.

Main Methods:

• Utilized the DDMoRe Interoperability Framework (IOF) with its Model Description Language (MDL) and R package.
• Developed and integrated a WinBUGS converter plugin for the IOF.
• Applied the IOF to a published diabetes study involving intravenous glucose tolerance test (IVGTT) data.
• Implemented pharmacostatistical models using various IOF target tools: NONMEM, WinBUGS, PsN, and Xpose.

Main Results:

• The WinBUGS plugin successfully integrated into the IOF, enabling complex Bayesian workflows.
• The IOF facilitated the execution of inter-related models for estimating insulin secretion rate from IVGTT data.
• Diverse uncertainty propagation approaches were implemented across different IOF target tools.
• The developed software supports extensive pharmacokinetic/pharmacodynamic (PK/PD) modeling features.

Conclusions:

• The DDMoRe IOF, enhanced with the WinBUGS plugin, provides a robust solution for interoperability and reproducibility in Bayesian pharmacometric modeling.
• The framework simplifies the encoding and execution of complex PK/PD models, particularly in WinBUGS.
• This approach facilitates the application of advanced modeling techniques to real-world disease studies, such as diabetes research.