Skip to menu

검색창 닫기

XEDITION

JoLab.AI :: Research

Main Hero Image
 

Genomics & AI

We combine large-scale genomic data with artificial intelligence (AI) to identify genetic variants linked to Alzheimer’s disease. By using machine learning techniques, we can analyze vast datasets to find key genetic markers, helping us identify individuals at higher risk more accurately.

Selected Publications

  • Uncertainty-aware genomic classification of Alzheimer's disease (Taeho Jo et al., Briefings in Bioinformatics, 2025) TrUE-Net framework combining transformer and random forest models with Monte Carlo dropout, achieving AUC 0.664. [LINK]
  • LD-informed deep learning for Alzheimer's gene loci detection using WGS data (Taeho Jo et al., Alzheimer's & Dementia: TRCI, 2025) Deep-Block framework analyzing 7,416 ADSP WGS participants, identifying novel SNPs within top 1,500 LD blocks. [LINK]
  • Deep learning-based identification of genetic variants (Taeho Jo et al., Briefings in Bioinformatics, 2022) SWAT-CNN approach, achieving AUC of 0.82. [LINK]
 

Neuroimaging & AI

Combining brain imaging technologies like PET and MRI with AI allows us to detect early changes in the brain associated with Alzheimer’s. This approach helps identify signs of the disease before symptoms appear, enabling timely interventions.

Selected Publications

  • Deep learning detection of informative features in tau PET for Alzheimer’s disease classification (Taeho Jo et al., BMC Bioinformatics, 2020) 3D CNN, 90.8% accuracy on tau PET scans. [LINK]
  • Deep Learning in Alzheimer’s Disease: Diagnostic Classification and Prognostic Prediction (Taeho Jo et al., Frontiers in Aging Neuroscience, 2019) Systematic review of deep learning with multimodal neuroimaging. [LINK]
 

Metabolomics / Proteomics & AI

By analyzing metabolites and proteins in blood and other biological fluids, we can track biochemical and molecular changes as Alzheimer’s progresses. Through integrated analysis of metabolomics and proteomics data using AI, we can predict the disease’s progression and customize treatment strategies even before cognitive decline begins.

Selected Publications

  • Longitudinal plasma proteomics: relation to incident Alzheimer's disease dementia and biomarkers (Eun Hye Lee, Taeho Jo, Kwangsik Nho, et al., Alzheimer's & Dementia, 2025) Seven-protein signature (ACES, C7, ZCD1, IL-17C, CC055, SO5A1) achieving AUC 0.848 for predicting incident ADD. [LINK]
  • Circular-SWAT for deep learning based diagnostic classification of Alzheimer’s disease (Taeho Jo et al., eBioMedicine, 2023) Serum-based metabolomics, accuracy up to 80.8%. [LINK]
 

Precision Medicine

By integrating genomics, neuroimaging, and metabolomics with AI, we can develop more precise predictions of disease progression. This personalized approach tailors treatments to individual patients, improving early detection and care.

Selected Publications

  • Uncertainty-aware genomic classification of Alzheimer's disease (Taeho Jo et al., Briefings in Bioinformatics, 2025) TrUE-Net framework for uncertainty-aware prediction, enabling identification of reliable predictions. [LINK]
  • Deep learning-based identification of genetic variants (Taeho Jo et al., Briefings in Bioinformatics, 2022) Highlighting APOE region; synergy with imaging/metabolomics. [LINK]
Up