A novel potential biomarker panel to diagnose depression derived from big proteomic data

Area of Science:

• Proteomics
• Biomarker Discovery
• Clinical Diagnostics

Background:

• Depression lacks a definitive clinical biomarker for diagnosis.
• The multifactorial nature of depression necessitates a panel of biomarkers rather than a single one.

Purpose of the Study:

• To identify and validate a panel of proteomic biomarkers for depression diagnosis.
• To assess the diagnostic accuracy of a novel biomarker panel for depression.

Main Methods:

• Analysis of two UK Biobank proteomic datasets (N=19,632 and N=19,374).
• Utilized Cox proportional hazards regression and machine learning algorithms (including LASSO).
• Validated biomarker panel performance using five-fold cross-validation and AUC.

Main Results:

• Identified 46 plasma proteins significantly associated with depression, primarily involved in immune pathways.
• A panel of six blood protein biomarkers achieved 75.4% diagnostic accuracy for depression.
• The six-biomarker panel's accuracy was comparable to using 46 or 2911 proteins.

Conclusions:

• Proteomic biomarkers show potential for early and population-based depression detection.
• The identified protein panel may serve as a valuable tool for depression screening and prediction.
• Further clinical and experimental validation is recommended for clinical application.