Structural signature of plasma proteins classifies the status of Alzheimer's disease - PubMed
5 hours ago
- #Alzheimer's disease
- #plasma proteins
- #machine learning
- Study profiles plasma protein structures from 520 participants, including AD, MCI, and healthy controls.
- Mass spectrometry and machine learning used to identify AD-specific structural proteome changes.
- Diagnostic panel developed using peptides from C1QA, CLUS, and ApoB for AD-associated structural changes.
- Three-marker panel achieved 83.44% accuracy in three-way classification (healthy vs. MCI vs. AD).
- Binary classification yielded AUCs of 0.9343 (healthy vs. MCI) and 0.9325 (MCI vs. AD).
- Longitudinal samples classified with 86.0% accuracy.
- Panel represents a promising diagnostic approach for early AD detection and clinical trial insights.