[Novel integrative multi-omics strategies of human's biological age computation.]

ABSTRACT

Multi-omics methods for analysing postgenomic data have become firmly established in the tools of molecular gerontology only in recent years, since previously there were no comprehensive integrative approaches adequate to the task of calculating biological age. This paper provides an overview of existing papers on multi-omics integrative approaches in calculating the biological age of a human. An analysis of the most common options for integrating methylomic, transcriptomic, proteomic, microbiomic and metabolomic datasets was carried out. We defined (1) concatenation (machine learning), in which models are developed using a concatenated data matrix, formed by combining multiple omics data sets; (2) fusion model approaches that create multiple intermediate submodels for different omics data to then build a final integrated model from the various intermediate submodels; and (3) transformation methods (via artificial intelligence) that first transform each of the single omics data sets into core plots or matrices, and then combine them all into one graph before building an integral complex model. It is unlikely that multi-omics approaches will find application in anti-aging personalized medicine, but they will undoubtedly deepen and expand the understanding of the fundamental processes standing behind the phenomenon of the biological aging clocks.