The 18th symposium

Speech

Special　Lecture1

“Epigenetics × AI-based Estimation of Biological Age”
Tsuyoshi Hachiya
CEO Genome Analytics Japan Inc.
Visiting Associate Professor
Department of Innovative Longevity Medicine, Juntendo University

Chronological age is an indicator representing the time elapsed since birth and is strongly correlated with the increasing risk of age-related diseases. However, even among individuals with the same chronological age, there are substantial differences in the degree of aging due to genetic variations and environmental factors, including lifestyle. Therefore, chronological age does not necessarily serve as a biomarker that accurately reflects the biological progression of aging in each individual, and the concept of “biological age” has been proposed to account for such inter-individual variability. Biological age is an indicator that quantifies the progression of aging based on physiological functions and molecular-level changes, and expresses it in units of age.
Biological age can be calculated based on physiological indicators such as cognitive function, muscular function, and immune function. It can also be estimated using comprehensive molecular measurements, including epigenomics, transcriptomics, and proteomics. In general, molecular and cellular changes are thought to occur prior to the decline in physiological function. Therefore, biological age estimated from molecular measurements has attracted attention as a biomarker capable of detecting aging and disease risk at an earlier stage.
A representative example is the epigenetic clock, which estimates biological age using artificial intelligence based on genome-wide measurements of epigenetic modifications such as DNA methylation. Epigenetic clocks are broadly classified into three categories: (i) models that predict chronological age, (ii) models that predict lifespan, and (iii) models that predict the pace of aging. These models have been increasingly applied in aging research and preventive medicine.
Most of the major epigenetic clocks proposed to date are AI models trained using cohort data from Western populations, and only a limited number of models have been constructed using data from Asian populations. Through analyses using Japanese cohort data, we have confirmed that models trained on Western population data can achieve comparable accuracy in predicting chronological age in Japanese individuals. However, for clinically more meaningful indicators—such as lifespan prediction and prediction of the pace of aging—it remains insufficiently validated whether these models have equivalent validity in Asian populations.
Because epigenetic states are strongly influenced not only by genetic factors but also by diet, lifestyle, and social environments, it is essential to construct models that take ancestral and regional differences into account.

Special　Lecture2

“Pioneering Anti-Aging Medicine with New Science”
Kazuhiro Sakurada
Professor, Department of Extended Intelligence for Medicine, Keio University School of Medicine & Graduate School of Medicine
Project Director, Predictive Medicine Special Project (PMSP), RIKEN Center for Integrative Medical Sciences
Specially Appointed Professor, Premium Research Institute for Human Metaverse Medicine (PRIMe), The University of Osaka

In his 1878 work On the Medical Path, Dr Shibasaburo Kitasato stated that 'the fundamental principle of the medical path is to prevent illness before it occurs'. Current medicine focuses on treating symptoms after injuries or illnesses have manifested. While these treatments relieve symptoms and extend lifespan, they do not actively improve health, leaving many people facing a low-quality life in old age. As we enter an ageing society, there is a need to develop proactive, accessible treatments that reduce the risk of age-related chronic diseases and extend the healthy lifespan of humans, improving quality of life even in old age. In this context, epigenetic reprogramming and the activation of pathways such as Nrf2 and DNA-PK are attracting attention. Aging progresses differently in each organ and varies from person to person. In the functional decline of organs associated with ageing, there are interactions between organs. We are pioneering new “AI for Medical Sciences” and “Organism Mechanics” to understand disease in the context of ageing and realize pre-emptive medicine based on high-precision prediction.