AI-Driven Analysis of Genetic and Transcriptomic Data in Alzheimer’s Disease
- PI
- Zhan, Huixin
- DAR ID
- 12088
- Initial Approval Date
- March 30, 2026
- Current DAR Status
- Expired
- PI Institution
- New Mexico Institute of Mining and Technology
- PI Institution Country
- United States
- PI Institution Type
- Non-Profit
Requested Datasets (4)
NG00067Alzheimer's Disease Sequencing Project Umbrella StudyNG00116Resolving mutations in challenging genomic regions to test association with disease phenotypesNG00174Whole Genome Sequencing (WGS) and SNP Array data for the SEA-AD Consortium CohortNG00075IGAP Rare Variant Summary Statistics- Kunkle et al. (2019)
Research Use Statement
Objectives:
This study aims to improve understanding of the genetic and molecular mechanisms underlying Alzheimer’s disease (AD) by applying advanced computational and deep learning models to existing genomic and transcriptomic datasets. Specifically, we seek to identify and characterize genetic variants associated with AD risk, progression, and related phenotypes, contributing to precision medicine approaches for neurodegenerative disorders.
Study Design:
This project involves secondary analysis of de-identified, controlled-access datasets from NIAGADS (NG00067, NG00116, NG00174, NG00027, NG00075). No new data will be collected. The data will be securely downloaded and analyzed on institutional servers at New Mexico Tech under an approved IRB and Data Use Agreement.
Analysis Plan:
We will integrate genomic, transcriptomic, and phenotypic data to develop and evaluate machine learning models—such as large language model–based architectures and disease-specific neural networks—to predict variant pathogenicity and gene-level associations. Phenotypic characteristics evaluated will include Alzheimer’s disease diagnosis, cognitive performance measures, neuropathological burden, and biomarker profiles (e.g., amyloid and tau levels). Statistical and model-based analyses will assess associations between genetic variants and these phenotypes, with results reported in aggregate, non-identifiable form.
Collaborations (if applicable):
N/A
This study aims to improve understanding of the genetic and molecular mechanisms underlying Alzheimer’s disease (AD) by applying advanced computational and deep learning models to existing genomic and transcriptomic datasets. Specifically, we seek to identify and characterize genetic variants associated with AD risk, progression, and related phenotypes, contributing to precision medicine approaches for neurodegenerative disorders.
Study Design:
This project involves secondary analysis of de-identified, controlled-access datasets from NIAGADS (NG00067, NG00116, NG00174, NG00027, NG00075). No new data will be collected. The data will be securely downloaded and analyzed on institutional servers at New Mexico Tech under an approved IRB and Data Use Agreement.
Analysis Plan:
We will integrate genomic, transcriptomic, and phenotypic data to develop and evaluate machine learning models—such as large language model–based architectures and disease-specific neural networks—to predict variant pathogenicity and gene-level associations. Phenotypic characteristics evaluated will include Alzheimer’s disease diagnosis, cognitive performance measures, neuropathological burden, and biomarker profiles (e.g., amyloid and tau levels). Statistical and model-based analyses will assess associations between genetic variants and these phenotypes, with results reported in aggregate, non-identifiable form.
Collaborations (if applicable):
N/A
Non-Technical Summary
This project uses advanced artificial intelligence and statistical tools to study the genetic and molecular factors that contribute to Alzheimer’s disease. By analyzing existing, de-identified research data from the National Institute on Aging’s NIAGADS repository, we aim to identify genetic variants and biological pathways linked to disease risk and progression. The study will combine information from DNA and gene-expression data to build computer models that can better predict how certain genetic changes affect brain health. Our ultimate goal is to improve scientific understanding of Alzheimer’s disease and support future efforts in early detection and personalized treatment.
Reported Publications
Publications shown are those reported by investigators through NIAGADS DAR reporting and renewal processes and may not represent a complete list of publications associated with the project.
No publications have been reported for this DAR.