Precision Oncology Beyond Genomics: The Integration of Multi-Omics and Artificial Intelligence

Abstract

Precision oncology has traditionally focused on genomic profiling to guide personalized cancer diagnosis and therapy. Although genomic medicine has significantly improved the identification of actionable mutations and targeted treatment strategies, genomic information alone cannot fully explain the remarkable biological complexity and clinical heterogeneity of cancer. Tumor behavior is influenced by multiple interconnected molecular layers, including epigenomics, transcriptomics, proteomics, metabolomics, lipidomics, microbiomics, and immunomics, which collectively regulate disease initiation, progression, therapeutic response, and resistance. The emergence of multi-omics technologies, combined with advances in artificial intelligence (AI), has enabled comprehensive characterization of cancer biology through integration of heterogeneous molecular and clinical datasets. Machine learning, deep learning, graph neural networks, transformer architectures, multimodal learning, and foundation models are increasingly capable of analyzing high-dimensional omics data alongside radiological imaging, digital pathology, electronic health records, and laboratory biomarkers. These computational approaches support improved cancer diagnosis, prognostic prediction, biomarker discovery, therapeutic selection, immunotherapy response prediction, drug discovery, and precision clinical decision-making. Despite remarkable progress, challenges remain regarding data harmonization, interoperability, computational complexity, explainability, ethical governance, regulatory validation, and equitable clinical implementation. This review provides a comprehensive overview of multi-omics integration and artificial intelligence in precision oncology, highlighting current applications, technological innovations, and future directions toward truly individualized cancer care.[1]