TY - JOUR
T1 - Network modeling and topology of aging
AU - Feng, Li
AU - Yang, Dengcheng
AU - Wu, Sinan
AU - Xue, Chengwen
AU - Sang, Mengmeng
AU - Liu, Xiang
AU - Che, Jincan
AU - Wu, Jie
AU - Gragnoli, Claudia
AU - Griffin, Christopher
AU - Wang, Chen
AU - Yau, Shing Tung
AU - Wu, Rongling
N1 - Publisher Copyright:
© 2024 Elsevier B.V.
PY - 2025/1/22
Y1 - 2025/1/22
N2 - Aging is a universal process of age-dependent physiological and functional declines that are strongly associated with human diseases. Despite extensive studies of the molecular causes of aging, little is known about the overall landscape of how aging proceeds and how it is related with intrinsic and extrinsic agents. Aging is a complex trait involving a large number of interdependent factors that change over spatiotemporal scales like a complex system. We develop an interdisciplinary form of statistical mechanics to reconstruct aging-related informative, dynamic, omnidirectional, and personalized networks (idopNetworks) from experimental or clinical data. The idopNetwork model can reveal how a specific biological entity, such as genes, proteins, or metabolites, mediates the antedependence of aging (i.e., the dependence of current trait values on their previous expression), identify how spatiotemporal crosstalk across different organs accelerate or decelerate the rate of aging, and predict how an individual's chronological age differs from his biological age. We implement GLMY homology theory to dissect the topological architecture and function of aging networks, identifying key subnetworks, surface holes and cubic voids that shape the rate of aging. Aging studies can be ideally conducted by monitoring molecular, physiological, and clinical traits over the full lifecycle. However, it is both impossible and ethically impermissible to collect the kind of data from which idopNetworks are reconstructed. To overcome this limitation, we integrate an allometric scaling law into the model to extract dynamics from snapshots of static data from a population-based cross-sectional study, expanding the utility of the model to a broader domain of cohort data. We show how this model can be used to unravel and predict the biological mechanisms underlying aging by analyzing an experimental metabolic data set of multiple brain regions in the aging mouse and a cross-sectional physiological data set of the lung for smoking and nonsmoking males aged from 20 years to nearly centenarians from the China Pulmonary Health Study. The model opens up a new horizon for studying how aging occurs through intrinsic and extrinsic interactions and could be used as a generic tool to disentangle human aging using various types of molecular, phenotypic or clinical data.
AB - Aging is a universal process of age-dependent physiological and functional declines that are strongly associated with human diseases. Despite extensive studies of the molecular causes of aging, little is known about the overall landscape of how aging proceeds and how it is related with intrinsic and extrinsic agents. Aging is a complex trait involving a large number of interdependent factors that change over spatiotemporal scales like a complex system. We develop an interdisciplinary form of statistical mechanics to reconstruct aging-related informative, dynamic, omnidirectional, and personalized networks (idopNetworks) from experimental or clinical data. The idopNetwork model can reveal how a specific biological entity, such as genes, proteins, or metabolites, mediates the antedependence of aging (i.e., the dependence of current trait values on their previous expression), identify how spatiotemporal crosstalk across different organs accelerate or decelerate the rate of aging, and predict how an individual's chronological age differs from his biological age. We implement GLMY homology theory to dissect the topological architecture and function of aging networks, identifying key subnetworks, surface holes and cubic voids that shape the rate of aging. Aging studies can be ideally conducted by monitoring molecular, physiological, and clinical traits over the full lifecycle. However, it is both impossible and ethically impermissible to collect the kind of data from which idopNetworks are reconstructed. To overcome this limitation, we integrate an allometric scaling law into the model to extract dynamics from snapshots of static data from a population-based cross-sectional study, expanding the utility of the model to a broader domain of cohort data. We show how this model can be used to unravel and predict the biological mechanisms underlying aging by analyzing an experimental metabolic data set of multiple brain regions in the aging mouse and a cross-sectional physiological data set of the lung for smoking and nonsmoking males aged from 20 years to nearly centenarians from the China Pulmonary Health Study. The model opens up a new horizon for studying how aging occurs through intrinsic and extrinsic interactions and could be used as a generic tool to disentangle human aging using various types of molecular, phenotypic or clinical data.
UR - http://www.scopus.com/inward/record.url?scp=85208277609&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85208277609&partnerID=8YFLogxK
U2 - 10.1016/j.physrep.2024.10.006
DO - 10.1016/j.physrep.2024.10.006
M3 - Review article
AN - SCOPUS:85208277609
SN - 0370-1573
VL - 1101
SP - 1
EP - 65
JO - Physics Reports
JF - Physics Reports
ER -
