[Omics and integrative methods for assessing biological age.]

Biological age, reflecting the extent of age-related changes in the organism at the molecular and systemic levels, serves as a more informative predictor of the risk of age-associated pathology and mortality compared to chronological age. This review systematizes current omics and integrative approaches to biological age assessment, analyzing their prognostic potential and limitations. Three generations of epigenetic clocks (Horvath, Hannum, PhenoAge, GrimAge, DunedinPACE), transcriptomic markers of cellular senescence (p16INK4a, mTOR, S100A9), as well as proteomic and metabolomic biomarkers, including indicators of inflammation (IL-6, TNF-α, С-RP), glycan clocks (GlycanAge), clinical-biochemical panels, and oxidative stress markers, are examined. It is shown that the integration of heterogeneous omics data (genomic, epigenomic, transcriptomic, proteomic, metabolomic) using machine learning algorithms (concatenation, ensemble methods) enables the creation of multi-omics models that outperform single-layer approaches in accuracy and allow for the assessment of aging heterogeneity across individual organs and systems (ageotype). Key limitations hindering the clinical implementation of these methods are discussed. Further progress in this field requires large-scale longitudinal studies, unification of methodological approaches, and validation of multi-omics models in independent cohorts to facilitate their integration into personalized preventive medicine algorithms.