GRPa-PRS: A risk stratification method to identify genetically-regulated pathways in polygenic diseases

Area of Science:

• Genetics
• Computational Biology
• Bioinformatics

Background:

• Polygenic risk scores (PRS) predict disease susceptibility but do not fully account for individuals who develop or evade disease.
• Unknown counteracting factors may influence disease outcomes despite genetic predisposition.
• Understanding these factors is crucial for advancing disease pathogenesis, prevention, and early intervention strategies.

Purpose of the Study:

• To develop a novel computational framework, GRPa-PRS, for identifying genetically-regulated pathways (GRPas).
• To stratify individuals based on PRS and clinical diagnosis to explore differential gene expression and pathways.
• To investigate disease risk and resilience pathways, and validate the framework's generalizability across polygenic diseases.

Main Methods:

• A PRS-based stratification framework was developed and applied to Alzheimer's disease (AD) cohorts.
• Genetically-regulated expression (GReX) was imputed, and differential GRPas were identified using enrichment and variational analyses.
• The framework was tested on AD and schizophrenia (SCZ) datasets, with and without APOE effects, and validated using orthogonality tests.

Main Results:

• Identified known AD-related pathways (e.g., amyloid-beta clearance, tau protein binding) and novel resilience pathways (e.g., calcium signaling).
• Discovered SCZ-associated pathways including mitochondrial function and muscle development.
• The GRPa-PRS method demonstrated greater consistency in identifying differential pathways compared to a variant-based PRS method.

Conclusions:

• The GRPa-PRS framework systematically explores differential GReX and GRPas across PRS strata.
• This approach provides new insights into pathways associated with disease risk and resilience.
• The framework is adaptable for studying other polygenic complex diseases.