Integrating Artificial Intelligence with Global Genomic Resources: A Narrative Review of Implications for Precision Medicine
Purpose This review aimed to explore how artificial intelligence can be integrated with global genomic resources and to examine its implications for advancing precision medicine, with particular attention to population diversity, predictive modeling, and clinical translation. Methods A narrative review design was adopted, drawing on literature from major databases and supplementary search sources including PubMed, Scopus, Web of Science, IEEE Xplore, Embase, and Google Scholar. Studies were selected based on relevance to artificial intelligence applications in genomic data analysis and precision medicine. Key information from included studies was extracted using a structured narrative extraction framework and synthesized thematically to identify key patterns and emerging insights. Results A total of 54 studies were included in this review. The synthesis identified five recurring application areas: genomic data integration, variant and disease association detection, disease susceptibility and risk prediction, treatment response prediction, and clinical decision support. Findings indicate that artificial intelligence can support the integration and analysis of multi-omics data, support the identification of genetic variants and disease associations, and improve predictive modeling for precision medicine. Incorporating diverse population data was also reported to improve model generalizability and reduce bias. However, challenges related to data standardization, interoperability, ethical governance, overfitting in small or biased cohorts, limited prospective clinical validation, reproducibility, and clinical implementation remain significant barriers. Conclusion The integration of artificial intelligence with global genomic resources holds substantial promise for advancing precision medicine by enabling more accurate, inclusive, and individualized precision medicine. However, this promise should be interpreted cautiously because many AI genomic models remain dependent on retrospective datasets, limited external validation, and variable reproducibility.