Hong Qin, PhD, MS | University of Tennessee at Chattanooga
Competition Sponsor: US National Academy of Medicine
Awardee Year: 2022
The molecular mechanism of the biological aging clock, such as the DNA methylation clock, remains obscure. We propose developing a knowledge-based neural network (KNN) model to infer the conserved mechanisms for longevity. We plan to first build our KNN using human genomics and phenotype data. We plan to use the multi-view technique to integrate heterogeneous genomic data, and use the DNA
methylation sites as a hidden layer of the neural network architecture. The main architecture of the neural network will be based on the hierarchy of functional components within cells. After the development of the neural network for human aging clocks, we will test the conserved mechanisms by applying transfer learning for aging clocks in mice and naked mole rats. The naked mole rats have exceptional longevity. By comparing trained KNN models in mice and naked mole rats, insights would be gained into how naked mole rats can achieve exceptional longevity.
To learn more about this proposal, email healthylongevity@nas.edu.