Identification of Mitochondrial Dysfunction Genes as Diagnostic Biomarkers for Ischemic Stroke by Integrated...

Area of Science:

• Biomolecular Sciences
• Genomics
• Computational Biology

Background:

• Current ischemic stroke (IS) diagnostics are limited in speed and accessibility.
• There is a critical need for advanced molecular diagnostic tools for IS.

Purpose of the Study:

• To develop a novel multi-gene diagnostic model for ischemic stroke (IS).
• To identify novel molecular biomarkers for improved IS diagnosis timeliness and accessibility.

Main Methods:

• Utilized three public gene expression datasets (GSE16561, GSE22255, GSE58294) for IS patients and controls.
• Integrated mitochondrial dysfunction genes with differentially expressed genes (DEGs) and Weighted Gene Co-expression Network Analysis (WGCNA) identified IS-related genes.
• Employed machine learning algorithms (LASSO, RF, SVM, NeighborMethod, NeuralNet, BayesMethod) for feature gene selection and diagnostic model construction.

Main Results:

• Identified 3548 DEGs and 100 mitochondrial dysfunction genes associated with IS.
• Discovered four key feature genes (MCL1, MRPL46, MTX3, RNASEH1) with significant diagnostic potential (AUC > 0.7).
• Developed machine learning models achieving high diagnostic accuracy (AUC range: 0.814-0.852) in the primary cohort and validated externally (AUC range: 0.730-0.783).

Conclusions:

• Established a robust multi-gene diagnostic model for ischemic stroke (IS).
• Identified novel molecular biomarkers that can enhance the speed and availability of IS diagnosis.
• The findings pave the way for more effective clinical management of IS.