Beyond One-Size-Fits-All: A Systematic Review of Genetic, Epigenetic, and Microbiome Contributions to Drug Action with AI Applications
Abdulrahman Abdulazeez, Arunkumar Jagadeesan, G. Muthukavitha, Aashish Choudhary · International Journal of Pharmaceutical Quality Assurance · 2025
Background: Drug response variability is influenced by multiple biological and computational factors, including pharmacogenomics, epigenetics, gut microbiota, and artificial intelligence (AI). Understanding these factors is crucial for optimizing personalized medicine approaches. While pharmacogenomics and epigenetics provide insights into genetic and environmental influences on drug metabolism, gut microbiota plays a pivotal role in modulating drug efficacy and toxicity. AI-driven models are revolutionizing drug response prediction by integrating these multifaceted variables into precision medicine frameworks. Objective: This systematic review synthesizes current evidence on how pharmacogenomics, epigenetics, gut microbiota, and AI collectively shape drug action, aiming to provide a comprehensive understanding of their roles in advancing personalized medicine. Methods: A systematic literature search was conducted across PubMed, Scopus, Web of Science, and Google Scholar, following PRISMA guidelines. Studies published between 2015 and 2023 focusing on the impact of pharmacogenomics, epigenetics, microbiota, and AI on drug response were included. Data extraction covered study characteristics, methodologies, and key findings, with meta-analysis performed where applicable. Bias risk was assessed using established quality evaluation tools. Results: From an initial pool of 1,279 studies, 40 met the inclusion criteria, with five eligible for meta-analysis. Pharmacogenomic variations were strongly linked to differential drug metabolism and adverse drug reactions, while epigenetic modifications influenced gene expression and drug response plasticity. Gut microbiota emerged as a key player in drug bioavailability, metabolism, and toxicity modulation. AI-driven algorithms, particularly machine learning models, demonstrated superior predictive accuracy in identifying drug response patterns and personalizing treatment regimens. Meta-analysis revealed a moderate overall effect size (SMD = 0.56, 95% CI: 0.29–0.83), with AI-driven models showing the highest impact on drug response predictions (SMD = 0.87, SE = 0.05). Conclusion: Pharmacogenomics, epigenetics, and gut microbiota significantly influence drug action, and AI offers a transformative tool to integrate these factors for precision medicine. The findings underscore the need for further research to validate AI-driven predictive models and to standardize methodologies for assessing drug response variability. Future studies should emphasize large-scale clinical trials, improved biomarker identification, and AI-powered decision-support systems to enhance therapeutic precision and patient outcomes.