A machine learning–enabled blood transcriptomic signature for digital diagnosis and subtyping of Alzheimer’s disease - Summary - MDSpire

A machine learning–enabled blood transcriptomic signature for digital diagnosis and subtyping of Alzheimer’s disease

  • By

  • Shuo Ma

  • Dawen Chen

  • Yanzhi Li

  • Yanxia Liu

  • Meiling Zhou

  • Jiwei Wang

  • Yuming Yao

  • Yinhao Chen

  • Guoqiu Wu

  • January 5, 2026

  • 0 min

Share

Objective:

To develop a non-invasive biomarker framework specifically for early detection and classification of Alzheimer's Disease (AD) using lactylation-related molecular signatures.

Key Findings:
  • The LDS model demonstrated high diagnostic performance, correlating with Braak stage and Mini-Mental State Examination (MMSE) scores, highlighting its significance compared to existing biomarkers.
  • LDS scores increased progressively from normal controls to aMCI and AD patients.
  • Combining LDS with plasma p-tau181 and p-tau217 improved detection of AD and amyloid-tau positive individuals.
Interpretation:

The study highlights the potential of lactylation as a novel biomarker for AD, linking metabolic and epigenetic alterations to disease progression.

Limitations:
  • The study primarily focuses on lactylation without exploring other potential biomarkers, which may limit the understanding of AD's complexity.
  • Further validation in larger, diverse cohorts is needed to confirm findings and enhance generalizability.
Conclusion:

LDS represents a promising, interpretable blood-based tool for early detection and stratification of Alzheimer's Disease, with significant potential implications for clinical practice.

Original Source(s)

Related Content