Higher criticism approach to detect rare variants using whole genome sequencing data

Area of Science:

• Genetics
• Statistical genetics
• Bioinformatics

Background:

• Single-variant tests in whole genome sequencing (WGS) data suffer from low statistical power.
• Association tests for variant groups are crucial for genetic mapping, especially for detecting sparse and weak genetic effects.
• The higher criticism (HC) approach is theoretically optimal for identifying such subtle genetic signals.

Purpose of the Study:

• To develop and assess a strategy for applying the higher criticism (HC) approach to whole genome sequencing (WGS) data.
• To evaluate the performance of HC in the context of rare variants, which are predominant in WGS data.
• To compare the effectiveness of the HC approach against other methods like the minimal p-value and sequence kernel association test.

Main Methods:

• Applied the higher criticism (HC) approach to Genetic Analysis Workshop 18 (GAW18) "dose" genetic data with simulated phenotypes.
• Investigated various strategies for grouping variants and collapsing rare variants within the HC framework.
• Compared HC performance with the minimal p-value method and the sequence kernel association test.

Main Results:

• The higher criticism (HC) approach demonstrated superior performance in detecting weak genetic effects.
• The strategy effectively handles WGS data characterized by a high proportion of rare variants.
• HC proved advantageous compared to the minimal p-value and sequence kernel association test in this context.

Conclusions:

• The higher criticism (HC) approach is a preferred method for detecting weak genetic effects in whole genome sequencing (WGS) data.
• The developed strategy successfully applies HC to WGS data, particularly for identifying associations involving rare variants.
• This work highlights the utility of HC for genetic mapping studies utilizing large-scale sequencing data.