Integrating multi-omics data for personalized nutrition using knowledge graphs and graph neural networks: A...

This comprehensive review focuses on the use of knowledge graphs and Graph Neural Networks (GNNs) to integrate multi-omics data in the field of personalized nutrition. It investigates how different biological datasets-like genomics, proteomics, metabolomics, and gut microbiome profiles-can be combined to create customized dietary plans. By reviewing scientific literature from 2015 to 2025, the study highlights how knowledge graphs help map complex biological interactions and how GNNs effectively analyze and interpret graph-based multi-omics information. The review demonstrates that knowledge graphs provide an elastic and scalable model for representing multi-relational biological information; however, GNNs enable the identification of complex non-local relationships between omics layers that cannot be detected using conventional statistics. The most significant benefits include improved forecast accuracy of individual responses to diet, enhanced biomarker discovery, and the potential for the development of customized nutritional therapy based on detailed characterization of the individual through comprehensive biological profiling. Nevertheless, several issues require further attention, including the quality and standardization of data across omics platforms, scalability uncertainty in computational models, interpretability shortcomings of models, and the ethical challenges associated with the use of personal genetic information. In this review, case studies are provided for comparing the graph-based approach with traditional methods and demonstrating how these methods outperform in terms of their capability to capture biological complexity. Future directions emphasize the importance of international cooperation, the utilization of technology in computational processes with potential for scalability, the development of models that are easier to interpret, and adherence to acceptable ethical standards: knowledge graphs and GNNs. KG and GNNs can couple multi-omics data, providing a paradigm shift towards precision nutrition by enabling unprecedented capabilities to optimize personal health outcomes through scientifically augmented, customized dietary interventions.