Variant-specific inflation factors for assessing population stratification at the phenotypic variance level

Area of Science:

• Genetics
• Epidemiology
• Bioinformatics

Background:

• Pooled analysis of participant-level data from multiple Whole Genome Sequencing (WGS) studies is common in genetic epidemiology.
• Differential variances in phenotypes across studies, termed 'variance stratification', can lead to reduced statistical power and increased false positive rates in pooled analyses.
• Existing methods may not adequately account for variance stratification, potentially compromising the reliability of genetic association findings.

Purpose of the Study:

• To investigate the impact of variance stratification on genetic association analyses using pooled WGS data.
• To develop a diagnostic procedure for identifying variance stratification in multi-study WGS analyses.
• To propose and implement a WGS-appropriate analysis approach that accounts for study-specific variances to improve performance.

Main Methods:

• Development of a procedure to compute variant-specific inflation factors for diagnosing variance stratification.
• Implementation of a WGS-appropriate analysis approach in freely-available software that accommodates study-specific variances.
• Illustration of the variance stratification problem and proposed solutions using simulations and real-world data from the Trans-Omics for Precision Medicine (TOPMed) Whole Genome Sequencing Program.

Main Results:

• Variance stratification was identified as a significant issue affecting pooled WGS genetic association studies.
• The proposed diagnostic procedure effectively identifies variant-specific inflation factors related to variance stratification.
• The WGS-appropriate analysis approach demonstrated improved performance by accounting for study-specific variances in simulations and TOPMed data.

Conclusions:

• Variance stratification is a critical consideration in pooled WGS epidemiological studies.
• The developed diagnostic tools and analysis methods provide a robust solution for addressing variance stratification.
• Accurate genetic association testing in large-scale WGS projects, like TOPMed, is enhanced by accounting for differential study variances.