Opportunities for machine learning and artificial intelligence in physiologically-based pharmacokinetic (PBPK)...

Area of Science:

• Pharmacokinetics and Drug Metabolism
• Computational Biology and Bioinformatics
• Pharmacology and Pharmaceutical Sciences

Background:

• Physiologically-based pharmacokinetic (PBPK) modeling is crucial for understanding drug behavior in living systems, especially when data collection is challenging.
• Advances have improved PBPK accuracy in special populations, increasing its value in drug development.
• Current PBPK models face limitations due to difficulties in defining complex biological mechanisms and parameter uncertainty.

Purpose of the Study:

• To review recent advancements in PBPK modeling influenced by machine learning (ML) and artificial intelligence (AI).
• To explore how ML/AI can address limitations in current PBPK models, such as parameter estimation and uncertainty quantification.
• To discuss future directions for ML/AI contributions to PBPK modeling in drug development.

Main Methods:

• Literature review of ML-influenced PBPK modeling advancements.
• Analysis of ML/AI applications in parameter estimation, model learning, database mining, and uncertainty quantification.
• Discussion of potential future integration of ML/AI into PBPK workflows.

Main Results:

• ML/AI tools show promise in improving parameter estimation and uncertainty quantification for PBPK models.
• These computational approaches can potentially overcome limitations of traditional PBPK modeling.
• ML/AI can facilitate earlier and more effective use of PBPK modeling in drug development.

Conclusions:

• ML and AI are poised to significantly enhance PBPK modeling capabilities.
• Integration of ML/AI can lead to more accurate and robust PBPK models.
• Future research should focus on leveraging ML/AI for broader applications of PBPK in pharmaceutical research and development.