A Zoom-Focus algorithm (ZFA) to locate the optimal testing region for rare variant association tests

Area of Science:

• Genomics
• Statistical Genetics
• Bioinformatics

Background:

• Analyzing rare variants in large-scale sequencing data is challenging due to low minor allele frequencies.
• Current methods often group functional and non-functional variants, potentially reducing statistical power.
• Genomic regions may contain numerous markers, not all relevant to the phenotype.

Purpose of the Study:

• To develop an algorithm to pinpoint optimal testing regions within larger genomic areas.
• To enhance the power of existing rare variant association tests.
• To improve the identification of disease-associated genetic markers from sequencing data.

Main Methods:

• The Zoom-Focus algorithm (ZFA) is proposed, acting as a wrapper for existing tests.
• ZFA involves a 'Zooming' step to find the best partition and a 'Focusing' step to refine boundaries.
• The algorithm is implemented as software available online.

Main Results:

• ZFA significantly increased the statistical power of various rare variant tests (SKAT, SKAT-O, burden, W-test).
• Application to real exome sequencing data for hypertensive disorder identified relevant metabolic disorder markers missed by gene-based methods.
• Simulation studies confirmed ZFA's effectiveness in boosting power for rare variant association studies.

Conclusions:

• ZFA is an efficient and powerful tool for enhancing association studies using whole exome or genome sequencing data.
• The algorithm effectively isolates relevant variants, improving the detection of genetic associations.
• ZFA offers a significant advancement for rare variant analysis in large-scale genetic studies.