Integrating transcriptomic and polygenic risk scores to enhance predictive accuracy for ischemic stroke subtypes

Area of Science:

• Genomics
• Computational Biology
• Cardiovascular Research

Background:

• Ischemic stroke (IS) presents complex etiological challenges globally.
• Genome-wide association studies (GWAS) and transcriptomic profiling are advancing IS risk prediction and mechanistic understanding.

Purpose of the Study:

• To integrate transcriptomic data with polygenic risk scores (PRS) for improved ischemic stroke (IS) risk prediction.
• To identify novel genetic susceptibility genes for IS and its subtypes.
• To evaluate the predictive performance of novel Sum Transcriptome-Polygenic Risk Scores (STPRS) models.

Main Methods:

• Transcriptome-Wide Association Studies (TWAS) were performed using GIGASTROKE and UK Biobank (UKB) data.
• Colocalization analysis identified shared variants between gene expression and IS risk.
• Sum Transcriptome-Polygenic Risk Scores (STPRS) were constructed and evaluated using logistic regression and AUC.

Main Results:

• TWAS identified 34 susceptibility genes for IS and its subtypes.
• Colocalization analysis implicated 18 genes in both gene expression and IS risk.
• STPRS models demonstrated superior predictive accuracy compared to PRS and showed associations with phenotypes like atrial fibrillation and blood pressure.

Conclusions:

• Integrating transcriptomic data with PRS via STPRS significantly enhances predictive accuracy for IS.
• This approach deepens the understanding of stroke's genetic architecture.
• STPRS offers a pathway for developing tailored stroke prevention and treatment strategies.