A recent study published in the journal Nature Aging investigated the genetic architecture of artificial intelligence (AI)-derived biological age gaps (BAGs) for multiple organ systems and their links with lifestyle, chronic disease, and aging.
Aging is a multifaceted biological process shaped by lifestyle, genetic, and environmental factors, which impacts organ systems and leads to chronic disease. Deciphering the phenotypic heterogeneity of aging across organs can lead to advances in precision medicine. One study investigated this heterogeneity using AI to estimate BAGs.
BAG represents the difference between individuals’ AI-predicted and chronological age. Notwithstanding the progress in multiorgan research, two questions remain: Which genetic variants influence the phenotypic heterogeneity of BAGs, and how are they causally linked to each other, lifestyle factors, and chronic diseases?
Study: Multiorgan biological age shows that no organ system is an island. Image Credit: Ws Studio1985 / Shutterstock
The study and findings
In the present study, researchers used computational genomics and AI to explore the genetic architecture of BAGs for nine organ systems and their causal links and associations with lifestyle factors, organ aging, and chronic diseases. They used multi-omics data of more than 370,000 participants from the United Kingdom (UK) Biobank.
First, support vector regression was used to derive BAGs for metabolic, musculoskeletal, cardiovascular, eye, brain, immune, renal, hepatic, and pulmonary systems using clinical and organ-specific imaging data. The BAGs were fit as phenotypes in a genome-wide association study (GWAS) to identify independent genetic signals, i.e., loci.
Subsequently, several downstream analyses were performed to validate the genetic signals. These included heritability estimation based on single nucleotide polymorphism, tissue-specific gene expression analysis, gene-set enrichment analysis, gene-drug-disease network analysis, causality analysis, and genetic correlation.
Overall, 393 BAG-genomic loci pairs linked to nine BAGs were identified. The researchers noted organ specificity as well as inter-organ crosstalk. The phenotypic and genetic correlations between BAGs were similar, supporting Cheverud’s conjecture (that phenotypic correlations likely are fair estimates of genetic correlations).
The tissue-specific gene expression analysis validated the genetic signals, which showed organ-specific enrichment. That is, genes related to cardiovascular BAGs were overexpressed or enriched in arterial and cardiac tissues. Further, the team identified potential causal associations among BAGs, lifestyle factors (e.g., body weight and sleep), and chronic diseases (e.g., diabetes and Alzheimer’s disease).
Conclusions
Together, the study reinforces that organ systems do not function in isolation, highlighting organ-specific correlations within organs and interlinks across organ systems. The cross-organ interlinks suggest that drugs for diseases in distinct organ systems could be repurposed, which could improve drug development success rates. The study’s limitations include the non-generalizability of the findings to diverse ethnic groups.
“We are highly excited about this study and its future research avenues. We foresee a paradigm shift from single-organ to multiorgan approaches, enabling more comprehensive modeling of human aging and disease.” – Junhao Wen, the lead author.
While there were strong correlations in GWAS beta values between European populations and other ancestral groups, studies should focus on under-represented groups. Moreover, sex differences were prominent in a few organ systems, especially the cardiovascular BAG. Overall, the results suggest that this research should be more comprehensively explored, jointly considering disease and aging, as sex differences often exist in chronic diseases, including autism and Alzheimer’s disease.
Researchers develop diabetes prediction system using clinical and genetic data