Revolutionizing drug discovery: Integrating artificial intelligence with quantitative systems pharmacology

Abstract

Quantitative systems pharmacology (QSP) provides a mechanistic framework for integrating diverse biological, physiological, and pharmacological data to predict drug interactions and clinical outcomes. Recent advances in artificial intelligence (AI) might transform QSP by enhancing model generation, parameter estimation, and predictive capabilities. AI-driven databases and cloud-based platforms might support QSP model development and facilitate QSP as a service (QSPaaS). However, challenges such as computational complexity, high dimensionality, explainability, data integration, and regulatory acceptance persist. This review critically evaluates the integration of AI within QSP, highlighting novel methodologies like surrogate modeling, virtual patient generation, and digital twin technologies. It also discusses current limitations and outlines strategies for future integration to enhance precision medicine, regulatory acceptability, and mechanistic interpretability in drug discovery and development.

Keywords: artificial intelligence (AI); digital twins; drug discovery; large language models (LLMs); machine learning (ML); quantitative systems pharmacology (QSP); regulatory acceptance.