Multi-Omics

Multi-omics is the integration of data across multiple omics layers, and compared with studies of a single omics type it offers the opportunity to understand the flow of information that underlies disease. ^[1] It usually refers to the crossover application of multiple high-throughput screening technologies such as genomics, transcriptomics, proteomics, and metabolomics. ^[2]

Multi-omics builds on high-throughput technologies including genotyping arrays that enabled genome-wide association studies, together with proteomics and metabolomics, and focuses on methods for integration across omics layers. ^[1] Data from omics sources such as genetics, proteomics, and metabolomics can be integrated with machine-learning predictive algorithms to unravel systems biology and discover new biomarkers. ^[3] Emerging directions extend integration to single-cell multi-omics and spatial multi-omics, combining multiple omics layers within single cells. ^[2]

Multi-omics deep phenotyping profiles the interaction of multiple levels of biology over time and empowers precision-health approaches, with applications that include longevity and aging. ^[4] Single-cell omics technologies, such as single-cell transcriptomics, enable high-dimensional profiling of individual cells and comprehensive characterization of brain aging at single-cell resolution. ^[5] Multi-omics analysis has been applied to aging alongside cancer, neurodegenerative diseases, and drug-target discovery. ^[2]

Integrative machine-learning methods applied to multi-omics enable discovery of biomarkers that can support accurate disease prediction, patient stratification, and delivery of precision medicine. ^[3] Multi-omics for precision health has been applied across genetic variation, cardio-metabolic diseases, cancer, infectious diseases, organ transplantation, and pregnancy, while clinical implementation still faces challenges. ^[4] Increasing biomedical data generation calls for multi-omics integration and interpretation by artificial-intelligence technologies, with machine learning central to biomarker discovery. ^[6]