Effective Visualizations Using "vachette" to Assess and Communicate Pharmacometric Model Results.

Area of Science:

• Pharmacometrics
• Computational Biology
• Data Visualization

Background:

• Effective communication of pharmacometric models is crucial for drug discovery and development decisions.
• Existing visualization methods may not adequately represent how models integrate diverse data and covariate effects.

Purpose of the Study:

• To introduce and describe the "vachette" visualization method.
• To demonstrate the utility and flexibility of the vachette method for pharmacometric models.

Main Methods:

• The vachette method uses user-provided model simulations and observations.
• It automatically generates a single plot overlaying observations onto a reference curve, accounting for covariates and random effects.
• Landmarks identify curve segments; transformations align segments, visualizing covariate effects and model fit.

Main Results:

• Vachette enables intuitive visualization of how models integrate data across subgroups and account for covariates.
• The method preserves the distance between model predictions and observations.
• Vachette-transformed data can enhance model assessments like visual predictive checks.

Conclusions:

• The vachette visualization method facilitates easier and more effective evaluation and communication of pharmacometric results.
• It is a valuable addition to the pharmacometrician's toolkit for informing critical decisions.
• The method's flexibility is demonstrated across various pharmacometric models.