Answers are grounded in this paper and cite their sources.
Background Pathological changes in Alzheimer's disease (AD) begin decades before clinical symptoms appear, highlighting the urgent need for non-invasive, reliable, and cost-effective blood-based biomarkers for early detection and staging. Objectives To identify plasma proteomic biomarkers for early diagnosis and staging of AD. Methods DIA-based high-resolution mass spectrometry was applied to plasma samples from 53 AD patients and 36 controls. A total of 2943 proteins with ≥70% valid values were quantified. Bioinformatics analyses identified dysregulated proteins and pathways, and machine learning models were trained and validated using cross-validation methods to construct diagnostic biomarker panels. Results AD patients showed downregulation of mitochondrial, nuclear, and plasma membrane proteins. Enrichment analysis highlighted alterations in cellular respiration and oxidative phosphorylation. Machine learning identified two biomarker panels: MBP (Myelin Basic Protein) /SOD2 (manganese superoxide dismutase) distinguished AD from controls (AUC=0.795; 70% accuracy, 58% specificity, 89% sensitivity), while MBP/PSMD10 (26S proteasome non-ATPase regulatory subunit 10) yielded an AUC of 0.996 for distinguishing mild-to-moderate AD within this strictly controlled pilot cohort (AUC=0.996; 94% accuracy, 100% specificity, 87% sensitivity). Conclusion The MBP/PSMD10 panel represents an exploratory, stage-specific candidate signature requiring rigorous cross-validation in heterogeneous, real-world prospective cohorts.
No takes yet. Share an insight, caveat, or question.
Han et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: