Poisson Approximation-Based Score Test for Detecting Association of Rare Variants

Area of Science:

• Genetics
• Statistical genetics
• Bioinformatics

Background:

• Genome-wide association studies (GWAS) excel at identifying common genetic variants but struggle with rare variants.
• The common disease rare variants (CDRV) hypothesis suggests rare variants contribute to common diseases, yet traditional GWAS methods lack power for their detection.
• Limited sample sizes in genetic studies further hinder the identification of rare variant associations.

Purpose of the Study:

• To develop a powerful statistical method for detecting associations between rare variants and common diseases.
• To address the limitations of traditional GWAS in identifying rare variants under the CDRV hypothesis.
• To provide an efficient approach for analyzing rare variants by pooling them based on functional relevance.

Main Methods:

• Proposed a novel statistical test, the Poisson Approximation-based Score Test (PAST), for rare variant association analysis.
• Developed two variants, ePAST and mPAST, employing different strategies for pooling rare variants.
• Utilized the Poisson distribution to approximate the distribution of total minor alleles for rare variants.

Main Results:

• Simulation studies demonstrated that PAST methods exhibit superior power compared to existing approaches.
• Application to the CRESCENDO cohort data confirmed the enhanced performance of PAST.
• The proposed methods effectively identify susceptible genes by pooling functionally related rare variants.

Conclusions:

• The Poisson Approximation-based Score Test (PAST) offers a more powerful approach for rare variant association analysis.
• PAST provides an effective solution for detecting rare variants contributing to common diseases, particularly in studies with limited sample sizes.
• The proposed methods advance the field of genetic association studies by improving the analysis of rare variants.