Artificial Intelligence Platform and Database for Taiwanese Healthy Aging

Project: National Science and Technology CouncilNational Science and Technology Council Academic Grants

Project Details

Abstract

This is a three-year proposal to create the first state-of-the-art Taiwanese Healthy Aging Artificial Intelligence (HAAI) platform and database. Machine learning will be applied to analyze the nationwide data of more than 100,000 people in the Taiwan Biobank, transforming health traits (lifestyle, behavior, biological measurements) and genetic data (epigenomic, genomic) into robust bias-free models for predicting the health trajectory of an aging Taiwanese individual compared to the rest of the population. Following my work published in NATURE 2017 (SCI, IF=41.577, R/C=1/64) and MOL PSYCHIATRY 2018 (SCI, IF=11.640, R/C=9/293), the current proposal aims at defining what is healthy aging in Taiwan and the factors behind age-acceleration. One key aim is to investigate the epigenome of the Taiwanese population that is highly suitable for the studying of aging because it retains information about the individual’s interaction with the environment. The data have been available in the Taiwan Biobank but remain under-valued and unpublished.Project aims include:1. Preparing high-quality input data, include identifying susceptible epigenetic and genetic loci from epigenome-wide association scans (EWASs) and genome-wide association scans (GWASs) of age-accelerated individuals for over 130 traits using biostatistics (e.g. lasso regression).2. Using GPU-accelerated H2O.ai machine learning algorithms (e.g. GLM, RF, GBM, DNN, XGBoost) to identify top variable importance from traits and genes (altogether processing 100 terabytes of data) that contribute to age-acceleration.3. Results from biostatistics and machine learning are aggregated to form prediction models for healthy aging classifications using deep neural networks.4. Establishing an online scalable database that can increase statistical power and accuracy with more individuals. The platform can determine whether a new individual has abnormal aging and provides a summary of the contributing variables using an intuitive web UI accessible to everyone.Findings from the HAAI platform and database are expected to provide new guidelines concerning healthy aging in the Taiwanese population. It aims to be the most comprehensive database in the Asia region. There are plans to expand internationally by collaborations with the United Kingdom databases and discover the factors that impact healthy aging worldwide.

Project IDs

Project ID:PB10907-4485
External Project ID:MOST109-2221-E182-045-MY3
StatusFinished
Effective start/end date01/08/2031/07/21

Fingerprint

Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.