Bridging the Worlds of Pharmacometrics and Machine Learning

Area of Science:

• Pharmacology and Computational Science

Background:

• Precision medicine necessitates individualized modeling of disease and drug dynamics.
• Machine learning (ML) is increasingly popular for computational modeling.
• A gap exists between pharmacometrics and ML, hindering progress.

Purpose of the Study:

• To bridge the gap between pharmacometrics and machine learning.
• To introduce pharmacometric problems and techniques to ML practitioners.
• To review ML approaches applicable to pharmacometrics.

Main Methods:

• Review of pharmacometric techniques including pharmacokinetic/pharmacodynamic (PK/PD) modeling, simulations, model-informed precision dosing, and systems pharmacology.
• Exploration of relevant machine learning approaches.
• Synthesis of how ML can address pharmacometric challenges.

Main Results:

• Identified key pharmacometric problems and techniques.
• Reviewed applicable machine learning methods.
• Highlighted the potential for synergistic effects between pharmacometrics and ML.

Conclusions:

• Facilitating collaboration between pharmacometricians and ML experts is crucial.
• Combining principled pharmacometric modeling with ML flexibility can enhance pharmacological applications.
• This integration promises synergistic advancements in precision medicine.